{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "# plt.style.use(\"ggplot\")\n",
    "plt.style.use(\"seaborn-v0_8-whitegrid\")\n",
    "\n",
    "import itertools\n",
    "from itertools import product\n",
    "import sys\n",
    "import pickle\n",
    "sys.path.append('../code')\n",
    "from classification import *\n",
    "from lp import *\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from scipy.spatial import ConvexHull"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Dataset\n",
    "* I preprocess the dataset of Obermeyer et al. (2019)\n",
    "* See `preprocess_health.py` for details\n",
    "* We can load the dataset by `df = pd.read_csv('../data/df_health_risk.csv')`\n",
    "\n",
    "* We make the outcome variable binary\n",
    "  - $\\tilde Y_i$ be the number of chronic diseases\n",
    "  - $Y_i \\coloneqq 1\\{\\tilde Y_i \\geq c_{97}\\}$, where $c_{97}$ be the 97-percentile of $(\\tilde Y_i)_i$.\n",
    "* Use all covariates except the group variable for prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 48784 entries, 0 to 48783\n",
      "Columns: 149 entries, trig_min-normal_tm1 to nt_bnp_min-high_tm1\n",
      "dtypes: float64(15), int64(134)\n",
      "memory usage: 55.5 MB\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('../data/df_health_risk.csv')\n",
    "Y_column = 'gagne_sum_t'\n",
    "G_column = 'dem_race_black'\n",
    "Y_columns = ['cost_t', 'gagne_sum_t']\n",
    "X_columns = list(set(df.columns) - set(Y_columns) - set([G_column]))\n",
    "X_columns.sort()\n",
    "\n",
    "y = df[Y_column].values\n",
    "thres_percentile = 97\n",
    "y_thres = np.percentile(y, thres_percentile)\n",
    "y_binary = (df['gagne_sum_t'].values > y_thres)*1\n",
    "df[Y_column] = y_binary\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.028779927845195147, 0.06288068792547474, 0.024373871580019444)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_b = df[df[G_column]==1]\n",
    "df_r = df[df[G_column]==0]\n",
    "df[Y_column].mean(), df_b[Y_column].mean(), df_r[Y_column].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.1144227615611676, 0.8855772384388324)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p_b = df[G_column].mean()\n",
    "p_w = 1 - p_b\n",
    "p_b, p_w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5582, 43202)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_b.shape[0], df_r.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Point estimate with different models\n",
    "* Figure 9 (a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.048, e_r=0.02, fairness=0.028\n",
      "[red] e_b=0.043, e_r=0.017, fairness=0.026\n",
      "r-skewed\n"
     ]
    }
   ],
   "source": [
    "clf_lr = LogisticRegression(max_iter=1000)\n",
    "print('Logistic Regression')\n",
    "res_logreg = CV_classification(df, Y_column, G_column, X_columns, verbose=True, clf=clf_lr, standardize=True, ratio=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.041, e_r=0.018, fairness=0.023\n",
      "[red] e_b=0.044, e_r=0.017, fairness=0.026\n",
      "r-skewed\n"
     ]
    }
   ],
   "source": [
    "clf_rf = RandomForestClassifier(n_estimators=500, random_state=1, n_jobs=-1)\n",
    "print('Random Forest')\n",
    "res_rf = CV_classification(df, Y_column, G_column, X_columns, verbose=True, clf=clf_rf, ratio=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlkAAAG7CAYAAADwnIBTAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAACBX0lEQVR4nO3dd1hT59sH8C+ICIJKcVbFWQGlIlqtYt2jigMRUetArWgd1S5brbW1rlptHVUc4KgDquJACs7+rK3aOqlWrQO3AhL2CAQyyHn/gOQlEnaS85yc+3NdXrYn5yR38vWEm+eMx4LjOA6EEEIIIcSgLPkugBBCCCHEHFGTRQghhBBiBNRkEUIIIYQYATVZhBBCCCFGQE0WIYQQQogRUJNFCCGEEGIE1GQRQgghhBgBNVmEEEIIIUZATRYhhBBCiBFY8V2APunp6di0aRPOnj2L1NRUtGjRApMmTYKfn1+5tj969Cj27NmDZ8+ewd7eHgMHDsQnn3yCOnXq6Kz366+/Yv78+XqfY+TIkVi1alWV3wshhBBCxIm5JksmkyEgIAAPHjzA+PHj0apVK5w6dQqLFi1CSkoKZs6cWer2wcHBWLduHTw9PfHFF18gLi4OISEhuH79OsLCwmBjY6NdNyYmBgCwYsUKWFtb6zxPs2bNDP/mCCGEECIazDVZoaGhuHPnDtatW4ehQ4cCAMaOHYvp06dj06ZNGDFiBF5//XW920okEgQGBqJXr14IDg6GpWXB0VA3NzfMmzcPISEhmD59unb9mJgY1K1bF6NHjzb+GyOEEEKIqDB3TlZERAQaNmyobbAAwMLCAtOmTYNSqURUVFSJ20ZFRUGpVGLKlCnaBgsAhg0bhiZNmiA8PFxn/ZiYGLRp08bwb4IQQgghosdUkyWVSvHkyRN06NCh2GOaZbdu3Spx+5s3b+qsW1T79u3x5MkTSKVSAEBqaiqSk5O1TZZCoYBCoajyeyCEEEIIARhrshITE8FxnN7Dgba2tqhTpw7i4uJK3F4ikaB27dqwt7cv9lijRo0AAPHx8QCA+/fvAwASEhLg6+sLDw8PuLu7w8/PD5cuXTLE2yGEEEKIiDF1TpZmlKlmzZp6H7exsUFubm6p25e2LVBwYj3w/ye9//PPP5g6dSrmzJmDZ8+eYefOnQgICMCmTZvQr1+/Ys+jUqmQmZmJGjVq6BySJIQQQgi71Go15HI56tSpAysr07Q/TDVZHMfp/K3v8bIam5K21ahWrRoAwN3dHTNnzoSvry+aN2+ufXzQoEEYNmwYli5dij59+hR7vczMTDx79qyst0IIIYQQBrVo0QJ169Y1yWsx1WTZ2dkBAPLy8vQ+npeXV+KVhZrt09PT9T6mGQHTHErs3LkzOnfuXGy9Jk2aYODAgfj111/x6NEjODs76zxeo0YNAAUh2dralvGOiLFJpVLUqlWL7zIIKAuWUBZsoTz4c+/ePWzYsAELFixAo0aN8OzZM+3PcVNgqslq2rQpLCwsIJFIij0mk8mQlZWlPbeqpO3v3LkDmUxW7LChRCKBpaUlGjZsWGYdmg43Jyen2GOakS1bW9sSD00S01EqlZQDIygLdlAWbKE8TE+lUiEsLAzDhg1DUFAQatasqT1dyJSn+jB1UpGdnR1at26N27dvF3tMc+Vgp06dStze3d0dgP4rEG/fvo02bdpoR7Jmz56NgQMH6h01e/z4MQC6IakQmOq4OikbZcEOyoItlIdpxcfHY8SIEbCwsECdOnV4bXCZarIAwNvbG/Hx8Th+/Lh2Gcdx2LlzJ6ytrTFkyJASt/Xy8kL16tWxY8cOnXOzjh07hpcvX8LX11e7rH79+njx4gXCwsJ0nuPKlSs4f/48evfubbJjtqTyNL+ZEP5RFuygLNhCeZjO3bt3oVarsX79eowfP57vcmDBlXWmuInl5eVh1KhReP78Ofz9/dGyZUucPHkSFy9exPz58xEQEAAAiI2NxfXr19GsWTN07NhRu/2mTZsQGBiI7t27w8vLC0+fPkVISAhcXV0RGhqqvcowJSUFo0ePRmJiInx8fODu7o5Hjx7hwIEDcHR0xL59+9C0adNi9clkMty7dw9t27al4V8GKJVKVK9ene8yCCgLllAWbKE8jE+pVGLRokWQyWTYuHGj3kOCfPz8Zq7JAoC0tDSsW7cOZ8+eRU5ODlq2bIkpU6bAx8dHu054eDgWLlyodyLn/fv3IzQ0FM+fP0e9evXQv39/fPTRR8UmiE5JScHGjRvx559/IjU1FY6OjujTpw/mzJlT4rlb1GSxJSEhodSLIYjpUBbsoCzYQnkYV05ODuLi4nDr1q1Sp8mjJksAqMkihBBC2BAZGYmgoCBERkaWee4bHz+/6Ww8E1CpVDRlj5EkJyejfv36fJdBINwsrK2tze7EZBo5YQvlYRxPnjzBH3/8gaNHjzK7D7NZlZngOA4vXrxAampqmTdJJZWXkpLCdwmkkBCzsLCwQN26ddGsWTNYWFjwXY5BlOdWNcR0KA/Devr0KebNm4d9+/Zh/fr1fJdTKmqyjCg1NRUpKSlo3LgxateubTZf4ISYC47jkJWVhZcvX8LOzg716tXjuySDSE5Oph/sDKE8DCchIQGzZ8/G1q1btReysYyaLCPhOA7x8fFwdHSkYWJCGGZnZ4e8vDzEx8ejbt26ZvHL0KsX+RB+UR5Vl5eXhyVLlmDRokU4duyYdoo81jF3nyxzoVKpoFKp8Nprr/FdCiGkDK+99pp2nzUHdF8mtlAeVZOWlgZvb2/06NEDtWrVEkyDBVCTZTRKpRIA6N4ohAiAZj/V7LdCx+pJwGJFeVTe0aNHYWlpiZCQEAwbNozvciqMmiwjM4dDD4SYO3PbT005NxspG+VRcXl5eZg1axauXbsGOzs7wZ7TRu01IYSYGblcrp2nlfCP8qiYR48eoVGjRhg/fjx69uzJdzlVQu01IYSYmVq1avFdAimC8ii/0NBQfPLJJ8jLyxN8gwVQk0WM5Msvv4SLi4vOH1dXV3Tt2hWTJk3C//73v2Lb9OvXD/369TNqXaZ4jcpKSkrSOUFW8xnGxcWZvJa4uDi4uLjgyy+/NPlrA4C/vz9cXFyq9BzPnz83UDXCk5qayncJpAjKo2wymQy5ublITk5GRESE2dxOhQ4XEqOaOXMmWrVqBaDgisu0tDScPHkSc+bMwcqVKzFq1CieK2RDREQEli5diqioKO10D2PHjoWnpyccHR15rs70Zs6cCT8/v0ptm52djQ8++ADNmjUrNq+pWDRq1IjvEkgRlEfp7ty5g48//hjbtm3Dp59+ync5BkVNFjGq7t27o2vXrjrLJk6ciMGDB2PdunXw9fU1u5OOK+Py5cvFLvPu2LEjOnbsyFNF/HrnnXcqvW1GRgb++ecfNGvWzIAVCYtEIqH78zGE8iiZSqXCTz/9hNDQULNsRulwITE5W1tbdOrUCSkpKUhLS+O7HELMjjn+sBIyyqO47OxsvP/++7h79y62b99utp8RNVkCl58P/PknsH9/wd/5+XxXVD5xcXFwcHCAg4NDqeu9ePECX3/9Nfr164c333wTHh4e8PHxwS+//FJs3ZSUFCxduhR9+/aFu7s7Bg4ciLVr1yI7O7vU11i+fDlcXFywaNEiqNXqMmuXSCT4+uuv0atXL7z55pvo1asXvv76a0gkEp31/P39MXDgQPz3338YP3483N3d0aNHDyxevFinuezXrx+OHj0KAOjfvz/8/f0BFD8n68qVK3BxccGxY8ewadMm9OvXD+3bt8fw4cPx22+/QaVSYevWrejXrx88PDwwcuRInD9/XqcmjuNw6NAhjB8/Hp07d4abmxt69OiBzz77rFLnMIWHh8PFxQXnz5/Ht99+i65du6JTp07w9/fH1atXq/TZFT0nS/M6V69exerVq9G7d2+8+eabGDx4MPbs2aOzXv/+/QEU3F/HxcUFV65cqfD7ErpXP0/CL8pDl1qthr+/P6ZMmQJ3d3e+yzEqOlwoYOHhwMcfA0XPi27aFNiwAfD15a+uoqRSqbahUKvVSE9Px+HDh3Hr1i0sW7as1Dv3xsXFwc/PD9bW1njvvffQsGFDJCUl4dChQ1i2bBlq166N4cOHAwASExMxatQopKenY/To0XB1dcXdu3exY8cO3LhxA3v27NH7WitWrEBoaCj8/f2xaNGiMg9dPn78GBMmTEB2djbGjBmDNm3a4MGDBzh06BB+//137Nu3Dy1bttSun5GRgSlTpqBz585YsGAB7t27h4MHD+LKlSsIDw+HnZ0dvvrqK+zatQvR0dFYuHAh2rRpU2oNa9asga2tLSZNmgSlUont27fj008/Rffu3fHy5UtMmjQJKpUK27dvx5w5c3DixAk0bdoUAPD9999jz549GDBggPbch+joaJw4cQJ3797FiRMnKnVPnyVLlgAAAgICoFarERoaivfffx9btmxB7969K/XZ6bNw4ULUrFkTkydPhpWVFfbt24eVK1fC3t4eo0aNQpcuXbBw4UJ8//336Ny5M8aMGYPWrVtX+P0IXd26dfkugRRBeRTgOA7bt29Hhw4dcPjwYUHdub2yqMkSqPBwwM8P4Djd5fHxBcsPH2aj0frwww/1Lh84cCBGjBhR6rYhISHIzMxEeHg43NzctMsHDRqEoUOH4vjx49oma926dUhOTsaOHTt0Lvt1cHBAcHAwLly4gD59+ug8/4oVKxASEoLp06fj888/L9f7WbZsGdLT07F79254enpqlw8YMABTp07F4sWLERISol2elZWFMWPGYPny5dplbdq0wcqVK7Fr1y7MmTMHAwYMwJkzZxAdHY0BAwZoG6KSqFQqHDp0SHvfHWtra6xcuRL379/HyZMntcvt7OywZMkSXLp0CaNHj0Z6ejp++eUX9O3bF5s3b9Y+34QJE6BWq3Hq1Cncu3dP57MuL4VCgaioKO00Uj4+PvDy8sKyZctw5swZWFhYVPiz08fe3h6HDh2CtbU1gIJ/R3379sWhQ4cwatQoODk5YcCAAfj+++/h5ORU5r8xcyWVSukHO0Moj4IGa9q0aWjTpg26dOkimhu0UpMlQPn5BSNYrzZYQMEyCwvgk0+AESMAvn9RWLBgAVxdXQEUjGRlZWXhn3/+QVhYGEaPHo09e/aUePXcl19+ienTp+tcyqtWq7Xzy2lOFOc4Dr///jtcXV2L3Vdl+vTp8PLyQosWLXSWf/fddwgJCYG/v3+5G6y0tDRcuXIFnp6eOk0CUHCitqenJy5duoTU1FSdL9SPPvpIZ93x48cjMDAQp0+fxpw5c8r12kX16tVL58aGb7zxBgCgT58+OsubN28OoGCUDyiYny86Ohr5rxxTzsrKgq2tLYDKz7E2adIknXk6GzVqhBEjRmD//v24e/cuXn/99Up9dq/y8vLSNlgA8Prrr6NevXpISUmpVN3mqkaNGnyXQIoQex43btyAhYUFVq5cKdg7t1cWNVkCdOGC7iHCV3EcEBtbsN4rgzcm5+bmVuzqwiFDhqBVq1ZYtmwZtm7dikWLFund1sLCAiqVCoGBgbh37x7i4uLw4sUL5ObmAihoroCCQ3JSqVTvYaFatWqhbdu2Osvi4+Oxd+9eWFpa4vr161CpVDpzi2VkZBSbw65mzZqIi4sDx3HapuZVbdq0waVLlxAXF6dtFF577TXUr19fZ73q1avDyckJT58+1fs8ZXn1+TS1v7pcMxTPFenGa9Sogd9//x1//PEHXrx4gbi4OCQkJGgPk5bnnDR9nJ2diy3T3LrjxYsXyM/Pr/Bnp4++e+dYW1tXum5zRZ8HW8Scx5YtW3D27FkEBweLcjRPHON1ZiYhwbDr8cHb2xsA9J4crXHp0iW8++672Lt3L6pVq4b+/ftj5cqVxU7m1oxsVeRWEF9++SVmzJiBO3fuICgoSOexuXPnokePHjp/fv75Z22zUtLraL5Ii460FP3vV2uu7PkIJU02W9b7VyqVmDVrFubMmYOYmBi0bt0aAQEB2LdvHz744INK1aKh731qRsysrKwq9dnpI5ZDDFWl2ScIG8SYR0ZGBmJjY+Hm5oZDhw6JssECaCRLkMp7uxWWb8ui+aFaWqOxePFi2NjY4Pjx4zqjNJrDXxqOjo6oWbOm3pGhlJQULFu2DEOHDsWgQYMAAE2aNMH7778PhUKB//3vfwgKCkL//v21I14LFixAVlaWzvM4OTnBzs4OAPDw4UO99T58+BAWFhY6w+HJycnIycnRbgsUnL8UFxenHekxlRMnTuDPP//EBx98gHnz5uk8prm6sbKePXuG7t276yx78uQJAKBFixbaL9iKfHak8jQ3tCVsEFse165dw4IFC7BmzRrthS9iRb8WClDPngVXEZY0cGFhATg5FazHKs0P9Vd/MBeVnp4OR0fHYoeItm/fDuD/fzusVq0a+vbtizt37uDatWs66x46dAinT5/WOwJibW2N7777Dvn5+ViwYAEUCgUA4M0330T37t11/jg5OcHR0RFvv/02Ll26hEuXLuk816VLl3DlyhW8/fbbOueYqdVq7Nq1S2fd3bt3QyaTYdiwYdplmvo4fSfaGUh6ejoAFJuu5vnz5zh9+jQAFDtfq7xCQ0Mhl8u1/x8XF4fIyEg4OzujTZs2lfrsKkvTuIv5EE1mZibfJZAixJIHx3HIyclBdHQ0jhw5gk6dOvFdEu9oJEuAqlUruE2Dn19BQ1X057Km8frpJ/5PegeAixcv6twjRqFQ4PLlyzh58iQaN26MgICAErft378/IiIi8OGHH6J3797Izc3Fb7/9huvXr8Pa2hpSqVS77ueff44rV65g6tSpGDduHFq3bo3bt28jPDwcffr00d476VUeHh6YNGkSdu/ejc2bN5c5pcO3336L8ePHY/r06Rg7dizeeOMNPHr0CGFhYXBwcMC3335bbJvt27fjxYsX8PDwwK1bt3D06FF06NABEyZM0K6jaSQ1V0cOGDCg1Doqo2fPnli7di2+++47vHjxAvXr18fDhw9x5MgRbcP66gheecXFxWH06NHw9fVFdnY2fvnlF+0VhRqV+ewq47XXXkO1atVw9epVHDx4EO+88w6aNGlikOcWilfPzyP8EkMeaWlpmDlzJiZOnIhZs2bxXQ4zqMkSKF/fgts06LtP1k8/sXH7BgDFzneytbVF48aNMXHiREyfPl3nirRXLV68GA4ODvjtt99w4cIFODo6wtnZGXv37kVYWBhOnDiBly9fonHjxmjcuDGOHDmCjRs34sSJE8jMzETTpk0xd+5cTJ06tdRzeT755BP8/vvv2L59O/r371/qzfHeeOMNhIeHY/PmzTh9+jTCwsJQv359+Pn5YdasWXoPd/3yyy9YsmQJTp48ifr162PmzJmYNWuWzvlH48aNw9WrV3HkyBFcvnzZKE1W69atsW3bNmzcuBE7d+4EUHB1nmaaIx8fH1y4cAGDBw+u8HMvWLAAd+/exaZNm2BhYYG3334bH330kc6oWWU+u8qwsbHB559/jm3btmH58uVYsmSJ6ObITExMpGlcGGLueXAchw0bNmD+/Pno3Lkz3+UwxYIz5vEJMySTyXDv3j20bdu21OPs5V2vqvLzC64iTEgoOAerZ082RrAItHc9j4mJ4bsUowkPD9fe/NOXlc6+Eky1vxJiTtRqNdauXQtHR8dSj0qwgo/9nEayBK5aNf5v00AIYUtCQoJZj5wIjbnm8c0336Bp06aYOnUq36Uwi5osQggxM/ruJ0b4Y255/PXXX3j48CGWL19Ot1UpA306hBBiZjIyMvgugRRhTnkEBQVh+/bt8PPzowarHGgkixAjKWsePnPg6+sr6HOxzBWdV8YWc8gjKSkJV65cga+vL2bMmFGhmz+LGbWhhBBiZsR4h3GWCT2Pc+fOYdy4cWjRogUaNGhADVYF0EgWIYSYGbponC1CzSM/Px8xMTGws7NDZGSkzswVpHxoJIsQQsxMjRo1+C6BFCHEPCQSCUaOHIl///0XnTt3pgarkmgkixBCzIxUKoWtrS3fZZBCQssjOzsbV65cwapVq9CuXTu+yxE0arIIIcTMGGIOSGI4QslDpVJhyZIlsLCwwPLly/kuxyzQ4UJCCDEzycnJfJdAihBCHhzH4eDBg2jXrh01WAZEI1mEEGJmzPHu4kLGeh6nTp3CyZMnsWHDBr5LMTs0kkUIIWYmISGB7xJIESznERERgaioKKxevZrvUswSjWQRQoiZadCgAd8lkCJYzOPFixfYuXMnvv32W/j4+PBdjtmikSxicF988QVcXFwQHR1d7LHs7Gy8+eabcHFxwbZt2/RuP3jwYLi7uyMvLw/h4eFwcXFBeHh4ma8bGBgIFxcXXLlyRWf58+fPK/dGeCKXyyGRSLT/X5HPwBj69euHfv368fLaJWVaEc+ePTNcQQKRmprKdwmkCNbyuHTpEqZNm4ZJkybR1DhGRp8uMbju3bsDAK5fv17sscuXL0OpVKJ69eq4cOFCscdTU1Px9OlTdO7cGTY2NhV63YEDB+KHH35A69attcu2b9+OoUOHVvAd8Ofu3bvw8vLCxYsXtcu6dOmCH374AV26dOGxMn7oy7QiVqxYgWnTphm4KvbVqlWL7xJIEazkoVAosHfvXrRr1w5RUVGV3q9I+dHhQmJwpTVZ58+fR82aNTFgwACcPHkS2dnZsLe31z5+7do1AMA777xT4dd1dXWFq6trsddTKpUVfi6+3L9/H/Hx8TrLnJyc4OTkxFNF/NKXaUWcPXvWgNUIR15enqDuy2TuWMjj+fPnmDFjBmbMmIE6derwWouY0EgWMbiGDRuiVatWuHHjRrHpJP766y+89dZb6NOnD5RKJS5fvqzzuOYQY2WaLEJIAToExBa+87h16xYsLCywdetWjBw5ktdaxIb2RKHKzATi4vQ/FhdX8DiPPD09kZGRgcePH2uXPX78GPHx8ejRowe6d+8OS0tLnD9/Xme7a9euoX79+sVGL3Jzc/HDDz+gT58+aN++Pby8vBASEqLTxL16/o6LiwuuXr2q/e8vv/xSu25iYiK++eYb9OrVC2+++Sb69u2LFStWID09vdzvUS6XY/PmzRg8eDDefPNNvP3225g5cyb+/fdfnfU0dd25cwefffYZ3nrrLXTp0gUzZ87E/fv3tet9+eWXWLhwIQBg4cKFcHFxAaD/nCwXFxcsW7YMx44dw4gRI9C+fXv07t0bW7ZsAcdxOHXqFHx8fODu7o6BAwdi586dxeq/cuUKZs2ahe7du8PNzQ1dunTB5MmTdQ5VlldcXBxcXFywZcsW7N69G/369YO7uzuGDx+OgwcPVvmz02SqeZ2dO3di//79GDZsGNq3b48ePXpgxYoVyM7O1lkvPj4e8fHxcHFxQWBgYIXfl1BZWdFBCpbwlYdcLsdHH32EkJAQODk5oWXLlrzUIWa0JwpRZiYweDCQlAT8+SdQ9FBSbCzQpw/QoAFw6hTA07Bw9+7d8csvv+D69et44403ABSMYgFAz5498dprr6Fdu3baZQCQlZWFBw8eYPjw4cWeb82aNWjSpAmmTJkCANi/fz9WrFgBa2trjB07Vm8NP/zwA4KCgvDkyRP88MMPaNasGQAgNjYW48aNg0KhwNixY9GkSRPcv38fBw4cwPnz53HgwIEy79Ccm5uLyZMn4+bNmxgwYAD8/f2RkpKCsLAwTJgwAWvWrIGXl5fONnPnzoWdnR3mzJmDrKws7N27F+PGjcMvv/yCdu3aYezYsbC2tkZYWBjGjh2Lt956q9Qa/vjjDxw7dgwTJ07EmDFjcODAAWzYsAG3bt3C9evXMXHiRIwdOxb79+/HDz/8gNdffx1DhgwBAPzvf//DRx99BFdXV0yfPh329vZ48OABDh8+jOnTp+O3335DkyZNSn19fQ4fPozU1FRMnDgRDRo0wLFjx/DNN98gPj4en376aaU/u1ft27cPOTk5GDduHF5//XWcPn0aISEhkEqlWL16NRwdHfHDDz/g+++/B6DbtIqBTCajueYYwkce2dnZeP78OQYOHKj3O5WYCEcqJCcnh4uOjuZycnIMsl6lxMZyXKtWHAcU/P3iRcHyFy90l8fGGv61yykrK4tr27Ytt2DBAu2ygIAArnfv3tr/X7duHefs7Mw9evSI4ziO+/333zlnZ2cuIiJCu86RI0c4Z2dnbujQoVxeXp52+YsXLzhnZ2du4sSJ2mUbN27knJ2ducuXL2uXTZw4kXN2dtapbfr06VynTp2458+f6yz/+++/OWdnZ+7bb78t8/1t2rSJc3Z25jZs2KCzPCkpievatSvXuXNnLisrS6eu4cOHczKZTLvunTt3OFdXV533oHm/R44cKXWZs7Mz5+zszN28eVO77N69e5yzszPn4uLC3bp1S7v80aNHnLOzM/f5559rl/n4+HDvvPNOsX+foaGhnLOzM/fzzz9rl/Xt25fr27dvqZ9HbGys9rWvXbumXS6Xyzk/Pz+ubdu22s+7Mp+dJlPN67Rv356Li4vTbpufn8+9++67nJubm85nXJ7aOc7I+ysPFAoF3yWQIkydR1hYGDd8+HBOpVKZ9HVZx8d+TocLhahp04IRrFatgCdPCkauLl4s+PvJk4Llf/5ZsB5PatWqBTc3N+3J73K5HNeuXdM516pHjx4ACi4nBgoOFVpYWOg9H2vo0KE6M9k7OTmhbt26SEpKqlBdWVlZuHDhAjp37gx7e3ukpaVp/7i6usLJyQn/+9//ynyeU6dOoWbNmpgxY4bO8vr162PSpEna1ylqxowZOie/tmvXDj169EB0dDTS0tIq9D6Ags/A3d1d+/+aK4WaN2+O9u3ba5c3b94cQMEhUo1Dhw4hMjISNWvW1C5TKBSwsLAAAOTk5FS4HqDgXLrOnTtr/9/a2hrvv/8+8vPz8fvvvwOo3Gf3qs6dO+uMtFlaWsLNzQ1KpRIZGRmVqt2cpKSk8F0CKcJUeXAch5iYGFy/fh1HjhxBtWrVTPK6pGR0uFConJwKGilNY6VpTDQNFgNXo3Xv3h1BQUFISUnBvXv3kJeXp22sAMDDwwN2dna4cuUKJk6ciOjoaLi4uKBevXrFnkvfMhsbmwpfOfjs2TOo1Wr8+eef8PT0LHG9vLw85OfnQyaT6SyvXr06HBwc8OLFCzRv3lyn8dNwdnYGUHBekL7lRbVq1Qrnz59HXFxchSeRffUzqV69OoCCZqUozfkgXJHz16ysrBAXF4ctW7bg6dOniIuLQ1xcHPLz84utWxElvUeg4OaHmr8r+tm9St+/B2trawDQvgcxY30aF7ExRR4xMTFYsGABDh48iFWrVhn99Uj5UJMlZE5OQEjI/zdYQMH/M9BgAQUnvwcFBeHff/9FdHQ0qlWrpr29A1DQFHTt2hVXr16FTCbD3bt3MXnyZL3PZajfyNRqNQBg0KBBeO+990pcz8rKCtu3b8emTZt0lr/99tvaE+41oz6v0vyQ1/zQ13j1/4uuW5kTYzVN1atKqquo4OBgrFu3Dk2aNEHnzp3RtWtXuLi4QKVSYfbs2RWuRaO096jJsDKf3av4vlqLdQkJCdRoMcTYebx48QLz5s1DcHBwmfsOMS1qsoQsNhbw99dd5u/PzEhWp06dYGNjg//++w9Xr15F+/bti92fpUePHjh79ixOnjwJlUqlM9JlDE0LD6HK5XKdhk/jzJkzcHBwgJWVFXx8fIqdfF67dm0AQLNmzfDixQvI5fJiIzIPHz4EADRu3Fhn+bNnz7SH7jSePHkCKysrk94HKyEhAevXr0eXLl3w888/63wpR0ZGVum59d1d/8mTJwCgvbKpMp8dqZiGDRvyXQIpwlh5yGQyfPPNN1i+fDkiIyPplw8GUSJCpbmKUHMO1t9/656jFRvLd4WwtrbGW2+9hcuXL+P+/ft6GyjNsrCwMNjY2JR5RV1Fab50NCNY9erVw1tvvYXz58/jn3/+0Vn3/Pnz+PDDD7XT/Tg5OaF79+46f958800ABSNhMpkMwcHBOs+RmpqK0NBQ2NnZFXu/u3bt0tYBAP/++y8uXryInj17au8I/Wq9xpCRkQGO49CqVSudBis3NxchISEAAJVKVannPnPmjM40NgqFAjt37oS1tTUGDBgAoHKfXWVZWloa9bNkVUXPVSTGZYw8kpOTMXz4cHh5eaFmzZrUYDGKRrKEKC6u+Enur56j1acPcO4crye/AwWHDNesWQMAen9wNm/eHE2bNsXNmzfRo0cPvefpVIXm3J2NGzfi7bffRvfu3fHtt99i4sSJmDJlCsaOHYs2bdrgyZMnOHDgABwcHLBgwYIynzcgIAB//PEHNm/ejAcPHsDT0xNpaWkICwtDVlYWfvjhB52TyoGCO+D7+/vDy8sLiYmJCA0NhYODA7766qti9UZGRoLjOKNM3PrGG2+gefPmOHLkCGrUqAFnZ2ckJSXh6NGjSE5OBgBIpdJKPbelpSXee+89TJw4Efb29oiIiMC9e/fw1VdfaX+br8xnV1n16tXDrVu3sGvXLnTs2BEeHh4GeV7WOTg48F0CKcLQeRw4cABeXl4ICwvTe34iYQc1WUJUq1bBfbAA3UODRRutBg0K1uOZ5pBc7dq1da6EK6pHjx44cOCAUQ4VTp8+HQ8ePMCOHTtw8+ZNdO/eXXtjzy1btuDUqVM4cOAA6tevj8GDB2P27NnFDunpU7NmTYSGhmLbtm04efIk/vzzT9SqVQtvvfUWpk2bpveH+erVq3HixAmsXbsWNWrUwMCBA/Hxxx/rXCXXrVs3DB8+HGfOnMHt27d1rtQzlOrVq2PHjh1Ys2YNjh8/joMHD6JBgwbo3LkzPvzwQ0yYMEHn/mUVMWTIELRp0wZ79uxBVlYWXF1dsXnzZu0oFlC5z66yPv74YyxevBhr166Ft7e3aJosmUxW4bk/ifEYKg+ZTIY5c+agRYsWGD16NF09KAAWXGUvIxIpmUyGe/fuoW3btqX+tl3e9SotMxOQSvWPVMXFFTRYND8VEwIDA7Fp0ybs3bsXXbt25bsco4iLi0P//v0xcuRIQV7ZZPT91cSkUikzkxITw+Rx7949ODk54b///kO3bt0MVJm48LGf00FcoapTp+RDgU2bUoNFiIiV5wpTYjpVyYPjOGzbtg0LFy6EUqmkBktgqMkihBAzI5fL+S6BFFHZPLKzsyGTyaBUKhEeHo7XXnvNwJURY6MmixBCzIy9vT3fJZAiKpPHjRs34O3tjdTUVHz44Yd09aBA0YnvhBjZ3LlzMXfuXL7LMKqmTZsiJiaG7zJIobS0NLoZKUMqkgfHcVCpVAgKCsLBgwfp6kGBo9aYEELMTKNGjfgugRRR3jwyMzMxYcIEPH78GMHBwdRgmQFqsgghxMxIJBK+SyBFlCeP/Px8+Pv746OPPoKrq6sJqiKmQIcLCSHEzNChQraUlgfHcdi4cSN69uyJiIgIOvfKzDCZZnp6OpYvX46+ffvC3d0d3t7eOHz4cLm3P3r0KHx8fODh4YEePXpg6dKlyMzMLHO7tWvXwsXFBVeuXKlK+TroNmSEsM/c9tOEhAS+SyBFlJQHx3HwL5x/tmPHjtRgmSHmRrJkMhkCAgLw4MEDjB8/Hq1atcKpU6ewaNEipKSkYObMmaVuHxwcjHXr1sHT0xNffPEF4uLiEBISguvXr2vnx9Pn6tWr2LFjh8HeR/Xq1QEASqXSYM9JCDEOzX6q2W+Frm7dunyXQIrQl8fly5dha2uLn376ic69MmPMNVmhoaG4c+cO1q1bh6FDhwIAxo4di+nTp2PTpk0YMWJEiUOvEokEgYGB6NWrF4KDg7W/Fbi5uWHevHkICQnB9OnTi22XlZWFBQsWwMrKCgqFwiDvw8rKClZWVkhPT6d5xAhhXHp6unafNQdZWVn0g5shr+axZs0a/PPPP9i6dSv9fDBzzH2jREREoGHDhtoGCyi4W+60adNw4cIFREVF4YMPPtC7bVRUFJRKJaZMmaIz7Dps2DCsW7cO4eHhepusJUuWQK1W47333sPevXsN8j4sLCzQpEkTPH/+HDY2NqhduzbdhZkQxnAch6ysLKSlpaF58+Zms4/SvIVs0eSRkpKC3NxcdOvWDfPmzTObf2+kZEw1WVKpFE+ePMHAgQOLPdahQwcAwK1bt0rc/ubNmzrrFtW+fXucOnWq2BxSEREROHHiBHbt2oXo6OiqvgUddevWRU5ODhISEvDy5UuDPjchxDAsLCxQr149szrEplar+S6BFKFWq3HhwgV8++232LhxI3r06MF3ScREmGqyEhMTwXGc3sOBtra2qFOnDuLi4krcXiKRoHbt2nrvrqu5T0l8fLz28tjY2FgsX74ckydPhqenp8GbLAsLCzRv3hxNmjQx2GFIoiszMxN1aJ5GJgg1C2tra7M5TKihUqn4LoEUUqvVyMjIwJ07dxAREYHatWvzXRIxIaa+WaRSKQCUODu2jY0NcnNzS92+tG2BghPrgYJ7ksyfPx+vv/46Pvvss6qUXSZzOteDNZaWlnRohBGUBTtK+h4kppWUlIQPPvgAU6dOLfOiLWKemLpeVHMZdUmXU3McV+YlrmVdil2tWjUAwNatW3H79m38+OOPqFGjRoVrVSgUkEgkUKvV2stzExISoFAokJKSgpycHGRmZiI9PR25ublISkpCfn6+zroqlQpJSUmQyWTIyMhAZmYmZDIZkpOToVQqddZVq9VITExEXl4e0tLSkJWVhezsbKSkpEChUOisy3EcEhISIJfLkZqaiuzsbEilUqSlpSEvL09v3UqlEsnJyRWqOzc3V1t3Tk5OhepOTU3V1q2pV/O3QqHQ1q05XyYvLw+JiYnF6k5LS9OpOyMjQ1u3SqXSWTc/P19bd3p6epl1SyQSbd1SqVRbt1wu16n31ezLU7fm85bJZNq6ZTJZherWl72mbrlcrq1bKpVq65ZIJGXWnZ6erq1bX/aaujMyMrR1JycnIy0trVjdiYmJ2rqzsrJKrJvjOG3dqampFao7JydHW3dubm6F69b3eWv+zWrqLm1f01d3Wlqatm4+viMSEhLoO6KEuk31HfHy5Uts2LABc+fORdeuXek7opR9zVTfEVlZWTA1C46hG8TExMTA29sbU6ZMwcKFC4s93qVLF7Rs2RIHDx7Uu72vry/i4uJw9erVYo+tXLkSe/bswYkTJ5CdnY3x48fj/fffx9SpU7Xr7NixAzt37sTmzZvRqVMnODg4FGvqZDIZ7t27h7Zt29JviwxQq9V0bxlGUBbsoCz4k5+fjxUrVqBly5aYNGkSAMqDFXz8/GYq9aZNm8LCwkLvFAQymQxZWVmlzgHVtGlT7W96r5JIJLC0tETDhg1x/vx5qFQqbN++HZ6ento/O3fuBAB8+OGH8PT0pJPVBSAxMZHvEkghyoIdlAV/Fi5ciObNm2sbLIDyEDOmThSys7ND69atcfv27WKPaa4c7NSpU4nbu7u74/Tp07h16xa6deum89jt27fRpk0b2Nvbw8fHB2+99Vax7SMiIvDrr79iwYIFcHV1Rf369av4joix0fQh7KAs2EFZmN6ZM2cQHx+PVatWFRu1ojzEi6mRLADw9vZGfHw8jh8/rl3GcRx27twJa2trDBkypMRtvby8UL16dezYsUPn3Kxjx47h5cuX8PX1BQA4OTmhe/fuxf44OTkBKLh5affu3St1rhYxLZo+hB2UBTsoC9Nav349Dh48iNGjR+s9LEh5iBdTI1kAMHnyZERGRmLBggX477//0LJlS5w8eRIXL17E/Pnz0aBBAwAFt1+4fv06mjVrho4dOwIAmjRpgpkzZyIwMBBTp06Fl5cXnj59ipCQELRv3x7vvfcen2+NGAHd1ZodlAU7KAvTiI+Px/Xr1+Hv71/qZ055iBdzI1k2NjYICQmBj48Pfv31V3z33XdIT0/H6tWrERAQoF3v2rVrmD9/PsLCwnS2nzNnDpYsWYKkpCQsW7YMJ0+exNixY7Fz5066vNwMZWRk8F0CKURZsIOyML7ffvsNkydPhrOzc5lNFOUhXkxdXSgEdHUhW2QyGeXACMqCHZSF8SiVSty/fx8KhQLt2rWDra1tmdtQHmwQ/dWFhFSUUqnkuwRSiLJgB2VhHC9evIC3tzcePnyIt956q1wNFkB5iBlz52QRUhE0EMsOyoIdlIXhZWVl4Z9//kFgYCDeeOONCm1LeYgXNVlE0KytrfkugRSiLNhBWRiOQqHAwoUL4ejoiEWLFlXqOSgP8aImiwhadnY2nevACMqCHZSFYXAch7CwMLzzzjvaWwBVBuUhXtRkEUFzdHTkuwRSiLJgB2VRdeHh4bhw4QLWr19f5eeiPMSLTnwngpacnMx3CaQQZcEOyqJqwsLC8Pfff2P16tUGeT7KQ7xoJIsIGk1XwQ7Kgh2UReU8evQIISEh+PbbbzF27FiDPS/lIV40kkUEjaarYAdlwQ7KouLOnTuHuXPnIiAgQO/UOFVBeYgXNVlE0DTTLBH+URbsoCzKLzc3F7t27YKHhwciIyPRrFkzg78G5SFe1GQRQUtNTeW7BFKIsmAHZVE+jx49gre3Nxo1aoQ6deqgevXqRnkdykO86JwsImi1atXiuwRSiLJgB2VRtuvXr6Nu3brYvXs3mjRpYtTXojzEi0ayiKDl5eXxXQIpRFmwg7IomUwmwwcffICjR4+iWbNmRm+wAMpDzGgkiwiaoU9QJZVHWbCDstAvMzMTz58/x+jRozFw4ECTvS7lIV7UZBFBs7Kif8KsoCzYQVkUt3v3bkRGRuLw4cNwd3c36WtTHuJF7TURtNzcXL5LIIUoC3ZQFv+P4zjcvn0bz58/x6FDh3gZVaI8xIuaLCJotWvX5rsEUoiyYAdlUeD27dsYOXIk2rZti2+//RbVqlXjpQ7KQ7yoySKCRpdGs4OyYAdlATx+/BjffPMNtm3bxvvhOspDvKjJIoJG01Wwg7Jgh5izyMrKwkcffYQmTZogPDyciRuBijkPsaMmiwgaTVfBDsqCHWLNIiEhAT4+Phg9ejRsbGyYuapPrHkQarKIwDVs2JDvEkghyoIdYsuC4zjs2bMH9vb2OHLkCHr27Ml3STrElgf5f9RkEUFLSkriuwRSiLJgh5iykEqlmDBhApKSkmBnZ4fXXnuN75KKEVMeRBfdvIMIGotfqGJFWbBDLFncunULrVq1wrx58/DWW2/xXU6JxJIHKY5Gsoig5eTk8F0CKURZsMPcs+A4Dhs2bMCKFSugVquZbrAA88+DlIxGsoigVa9ene8SSCHKgh3mnEVWVhYsLCxQs2ZNhIWFwcLCgu+SymTOeZDS0UgWIYQQQbh8+TJGjBgBqVSK6dOnC6LBIuJGI1lE0BQKBd8lkEKUBTvMLQuO4yCXy7Fnzx4cPXoUDg4OfJdUIeaWByk/GskigmZvb893CaQQZcEOc8oiJSUFfn5+iI+Px9atWwXXYAHmlQepGGqyiKClp6fzXQIpRFmww1yyUKlUeP/997Fo0SK0bt2a73IqzVzyIBVHhwuJoNFN/thBWbBD6Fmo1Wr88MMPGDp0KCIjIwV/7pXQ8yCVRyNZRNAkEgnfJZBClAU7hJyFWq3G2LFj4ejoiDfffFPwDRYg7DxI1dBIFhE0mniVHZQFO4Saxblz51C3bl0EBwfD0dGR73IMRqh5kKqjkSwiaDTxKjsoC3YILQuO47B8+XLs3r0bLVq0MKsGCxBeHsRwaCSLCFrdunX5LoEUoizYIaQsJBIJVCoVBg4ciK5du5rF4cFXCSkPYlg0kkUELSsri+8SSCHKgh1CyeLMmTOYMGECcnJy0K1bN7NssADh5EEMj0ayiKDZ2NjwXQIpRFmwg/UsVCoVcnNz8eTJE0RFRaFmzZp8l2RUrOdBjIdGsoigqdVqvksghSgLdrCcRXx8PHx8fBAdHY0PPvjA7BssgO08iHHRSBYRtPz8fL5LIIUoC3awmgXHcdi6dSvWrFkDV1dXvssxGVbzIMZHTRYRNFtbW75LIIUoC3awloVSqcTixYvRoUMHrFixgu9yTI61PIjp0OFCImiZmZl8l0AKURbsYC2L+fPno0OHDnjvvff4LoUXrOVBTIdGsoig1atXj+8SSCHKgh2sZHHs2DFkZGRg3bp1ZnvlYHmwkgcxPRrJIoKWlJTEdwmkEGXBDhayWLVqFU6fPg0/Pz9RN1gAG3kQftBIFhE0mq6CHZQFO/jM4tmzZ7h9+zY++OADs7tze2XRviFeNJJFBI2mq2AHZcEOvrKIjIzEjBkz4ObmRg1WEbRviBeNZBFBq1+/Pt8lkEKUBTtMnYVcLkdMTAycnJwQFRUFa2trk74+62jfEC8aySKClp6ezncJpBBlwQ5TZvH48WMMHz4csbGx6NixIzVYetC+IV40kkUEzc7Oju8SSCHKgh2myiIjIwO3bt3C9u3b0bx5c5O8phDRviFe1GQRQVMoFKKYlkMIKAt2GDuLvLw8fPbZZ2jVqhU+//xzo72OuaB9Q7yoySKEEFJuHMdh//79GDZsGIYMGcJ3OYQwjZosImh0/gc7KAt2GCuLffv24d9//8UPP/xglOc3V7RviBed+E4ELScnh+8SSCHKgh3GyGLv3r24c+cOVq5cafDnNne0b4gXjWQRQXvttdf4LoEUoizYYcgs7t27hwMHDmDJkiWiv3N7ZdG+IV40kkUELTk5me8SSCHKgh2GyuJ///sf5s+fj5kzZ1KDVQW0b4gXjWQRQaPpKthBWbCjqlnk5OTgwIEDGD16NCIiIlCtWjUDVSZOtG+IF41kEUGj6SrYQVmwoypZ3Lt3D8OHD0fLli1Ru3ZtarAMgPYN8aKRLCJoDRo04LsEUoiyYEdlsuA4DteuXUODBg2wf/9+NGzY0AiViRPtG+JFI1lE0FJSUvgugRSiLNhR0SykUikmT56MM2fOoEWLFtRgGRjtG+JFI1lE0GrXrs13CaQQZcGOimSRnp6OZ8+eYdq0aejVq5cRqxIv2jfEi5osImh5eXmwtbXluwwCyoIl5cmC4zgEBQXh3Llz2L9/P109aES0b4gXHS4kgkYn5bKDsmBHWVlwHIcbN25AKpVi37591GAZGe0b4sVkk5Weno7ly5ejb9++cHd3h7e3Nw4fPlzu7Y8ePQofHx94eHigR48eWLp0KTIzM4utl52djbVr12LQoEFwd3fHoEGDEBwcDKVSaci3Q4yIvrzYQVmwo7QsoqOjMXLkSHTo0AHz58+HpSWTPwbMCu0b4sXc4UKZTIaAgAA8ePAA48ePR6tWrXDq1CksWrQIKSkpmDlzZqnbBwcHY926dfD09MQXX3yBuLg4hISE4Pr16wgLC4ONjQ0AQKlUYsaMGbhx4wb8/PzQrl073LhxA+vWrcP169cRHBxsirdLqig3Nxd2dnZ8l0FAWbCkpCzu37+P1atXY+fOnfSD34Ro3xAxjjHBwcGcs7Mzd+zYMe0ytVrNBQQEcG5ubtzLly9L3DYhIYFzc3Pjpk2bxuXn52uXR0VFcc7Ozty2bdu0yw4fPlxsGcdx3PLlyzlnZ2fu77//1vsaOTk5XHR0NJeTk1PZt0gMSC6X810CKURZsOPVLNLS0rjZs2dzcrmcU6vVPFUlXrRvsIGPn9/MjRNHRESgYcOGGDp0qHaZhYUFpk2bBqVSiaioqBK3jYqKglKpxJQpU3SGwIcNG4YmTZogPDxcu0wqlcLZ2RmjR4/WeY533nkHAHDnzh1DvSViRKmpqXyXQApRFuwomkVsbCx8fX0xefJkWFtb0/lXPKB9Q7yYarKkUimePHmCDh06FHtMs+zWrVslbn/z5k2ddYtq3749njx5AqlUCgCYMmUKoqKi4ODgoLOeprlq0qRJpd4DMS2aroIdlAU7Xn/9dXAch+3bt8PR0RERERF4++23+S5LtGjfEC+mmqzExERwHKf3H6StrS3q1KmDuLi4EreXSCSoXbs27O3tiz3WqFEjAEB8fHyxx+RyOR4/fozg4GAEBQXB3d0dAwYMqMI7IaZC01Wwg7JgR0xMDEaPHg25XI6aNWuiTp06fJckarRviBdTJ75rRplq1qyp93EbGxvk5uaWun1p2wIFJ9a/KiwsDN999x0AwNHREUuWLIG1tXWFaif80DTPhH+UBRuuX7+ONm3aYPHixXB3d+e7HALaN8SMqZEsjuN0/tb3eFmXG5e0rYa+K2o8PDywefNmfPPNN6hRowbGjh2LP/74o9TnUSgUkEgkUKvV2t9SEhISoFAokJKSgpycHGRmZiI9PR25ublISkpCfn6+zroqlQpJSUmQyWTIyMhAZmYmZDIZkpOToVQqddZVq9VITExEXl4e0tLSkJWVhezsbKSkpEChUOisy3EcEhISIJfLkZqaiuzsbEilUqSlpSEvL09v3UqlEsnJyRWqOzc3V1t3Tk5OhepOTU3V1q2pV/O3QqHQ1p2VlaWtOzExsVjdL1++1Kk7IyNDW7dKpdJZNz8/X1t3enp6mXVLJBJt3VKpVFu3XC7XqffV7MtTt+bzlslk2rplMlmF6taXvaZuuVyurVsqlWrrlkgkZdadnp6urVtf9pq6MzIytHUnJyfj5cuXxepOTEzU1p2VlVVi3RzHaetOTU2tUN05OTnaunNzcytct77PW/NvVlN3afuavrrT0tK0dZvqO0KlUuHrr7/Gjz/+iOfPn6NFixb0HaGnbj6+IxITE+k7opR9zVTfEVlZWTA1C66srsSEYmJi4O3tjSlTpmDhwoXFHu/SpQtatmyJgwcP6t3e19cXcXFxuHr1arHHVq5ciT179uDEiRNo3bp1iTVIJBIMHToUderUwdmzZ4s9LpPJcO/ePbRt27bEUTNiOnK5HDVq1OC7DALKgk/p6emwtLREREQEJk2aBIVCQVkwhPYNNvDx85upkaymTZvCwsICEomk2GMymQxZWVmlDrs2bdpU+5veqyQSCSwtLcuc+LRRo0bo0qUL4uPjkZ6eXvE3QUwqOzub7xJIIcqCH+fOncOoUaMgl8sxefJkWFhYUBaMoTzEi6kmy87ODq1bt8bt27eLPaa5crBTp04lbq85/0DfFYi3b99GmzZttCfFz5gxA/369YNKpSq2bk5ODiwtLem8LAGgjNhBWZgWx3GQyWQ4fPgwfv31VzRo0ED7GGXBFspDvJhqsgDA29sb8fHxOH78uHYZx3HYuXMnrK2tMWTIkBK39fLyQvXq1bFjxw6dc7OOHTuGly9fwtfXV7uscePGiI+Px5EjR3SeIzo6GtHR0ejevTvdoZcQwiSJRAIfHx8kJycjMDAQtWrV4rskQogeTF1dCACTJ09GZGQkFixYgP/++w8tW7bEyZMncfHiRcyfP1/721psbCyuX7+OZs2aoWPHjgAK7m01c+ZMBAYGYurUqfDy8sLTp08REhKC9u3b47333tO+zpw5c3D+/HksX74cd+/eRbt27fDw4UOEhYXB0dERixcv5uX9k4pRKBR8l0AKURamoVAoMH36dKxatQrNmzcvcR3CDspDvJg68V0jLS0N69atw9mzZ5GTk4OWLVtiypQp8PHx0a4THh6OhQsXYuTIkVi1apXO9vv370doaCieP3+OevXqoX///vjoo4+K3SsmLS0NGzZswNmzZ5GWloa6deuiT58++PDDD0s8d4tOfGcLnVDKDsrCuPLz87Fs2TKMGzcOLi4upd65nbJgC+XBBj5+fjPZZLGMmiy2SCQSugcNIygL48nPz8eoUaMwatQo+Pv7l7k+ZcEWyoMNfPz8Zu5wISEVUdbVosR0KAvjOH36NJycnLB79+5i04CVhLJgC+UhXsyd+E5IRei73QfhB2VhWBzHYeHChTh69ChatmxZ7gYLoCxYQ3mIF41kEUGjiVfZQVkYTmxsLCwtLTFq1Ch07ty5wttTFmyhPMSLRrKIoNHEq+ygLAzj+PHjmDp1KuRyeaUaLICyYA3lIV40kkUErW7dunyXQApRFlWjVCqRl5eHhIQEREVFaSe1rwzKgi2Uh3jRSBYRND4m/CT6URaV9+zZMwwfPhw3b97EtGnTqtRgAZQFaygP8aKRLCJotra2fJdAClEWlcNxHLZt24YtW7agVatWBnlOyoItlId4UZNFBC0/P5/vEkghyqJi5HI5FixYgF69emHlypUGfW7Kgi2Uh3jR4UIiaPTlxQ7KomI+//xz9O3bV2dOVUOhLNhCeYgXjWQRQavquSvEcCiL8jl8+DCUSiU2btxY6tQ4VUFZsIXyEC8aySKCRieUsoOyKNuSJUtw9epV+Pn5Ga3BAigL1lAe4kUjWUTQ6tWrx3cJpBBlUbKHDx8iJiYGn3zySYXu3F5ZlAVbKA/xopEsImhJSUl8l0AKURb6HTp0CB9//DE8PDxM0mABlAVrKA/xopEsImg0XQU7KAtdubm5iImJQZs2bRAZGQkrK9N93VIWbKE8xItGsoig0XQV7KAs/t+9e/cwfPhwJCcnw8PDw6QNFkBZsIbyEC8aySKCVr9+fb5LIIUoiwJpaWm4f/8+9u7di8aNG/NSA2XBFspDvGgkiwhaeno63yWQQmLPIicnBwEBAdi3bx9GjhzJW4MFUBasoTzEi0ayiKDZ2dnxXQIpJOYsOI7D/v37MX78ePTv35/vckSdBYsoD/GqcpOVn5+PatWqGaIWQipMoVCgZs2afJdBIM4sOI7Dzz//jEePHuH777/nuxwtMWbBMspDvKp8uLB3797YsmULUlNTDVEPIYQIxs8//4yEhASsWLGC71IIIQyq8kjWd999hwMHDmDbtm0YOHAg/P394e7ubojaCCmTtbU13yWQQmLK4ubNm4iIiMDixYuNeuf2yhJTFkJAeYiXQUaytm7dihMnTqBx48aYPXs2/Pz8EBkZCaVSaYgaCSlRTk4O3yWQQmLJ4tixY1i6dClmz57NZIMFiCcLoaA8xMuC4zjOEE+kUCgglUqRmpqK8PBwhIWFwc7ODn/99Zchnp4ZMpkM9+7dQ9u2bekYOwNUKpXJ70FE9DP3LLKysnDw4EGMGTMG9vb2sLRk9+Jsc89CaCgPNvDx87vKqffo0QNSqRT5+fmws7PT/nF1dYW9vb0haiSkRMnJyXQ3ZUaYcxa3bt3Cxx9/jBUrVqB27dp8l1Mmc85CiCgP8apyk+Xk5ITY2FhMnjwZY8aMQZ06dQxRFyHlQl9c7DDHLDiOw+XLl/H666/j8OHDqFu3Lt8llYs5ZiFklId4VXm8e//+/di2bRseP36Md999F9988w0ePnxoiNoIKRNNV8EOc8siPT0d48aNw+XLl9GiRQvBNFiA+WUhdJSHeBnsnCygYDqJ/fv3IywsDC1btoS/vz8GDBhgqKdnAp2TxRa6Txs7zCmL1NRUPHv2DCqVCl27duW7nAozpyzMAeXBBkGek3X06FHk5ORAJpMhJycHOTk5ePvtt/H7779j7ty5uHfvniHqJESvlJQUNGzYkO8yCMwjC47jsH79ety8eRO7d+9m9urBsphDFuaE8hCvKjdZu3fvRu3atVGrVi3t382bN8fHH39M52cRo6N/Y+wQehZqtRrXrl1D9erVBd1gAcLPwtxQHuJVqSZLJpOB4zjY2dnh119/NXRNhJRbbm4ubGxs+C6DQNhZXLx4EWvWrMHhw4cFeXjwVULOwhxRHuJVoSbr1KlT2LBhA549ewYAsLe3x7Vr14xRFyHlQuc5sEOoWdy6dQtbtmzBrl27mL73VUUINQtzRXmIV7m/Uc6cOYNPPvkEqampGDx4MNq2bYvs7Gzt4zdv3sQ333yjbcAIMQVz+aFoDoSWRXJyMmbNmoV27dohJCTErA7pCC0Lc0d5iFe5kw8KCoKDgwOioqKwfv169OvXT+fxtm3b4rfffsPRo0cNXiQhJcnLy+O7BFJISFk8ffoUY8eOxYwZM2BlZSXo86/0EVIWYkB5iFe5m6wHDx5g8ODBJV4hYW1tjc6dO+PixYsGK46Qsgjh7ttiIYQs8vPzsWXLFjRs2BCRkZHw8PDguySjEEIWYkJ5iFe5myxra2vUqFGj1HUaNWpEN10jJpWamsp3CaQQ61mkpqbC19cXNWrUgK2trVlP+8V6FmJDeYhXuU98d3FxQXR0dKnr2NraIiMjo6o1EVJuNF0FO1jO4sqVK3Bzc8OqVavQtm1bvssxOpazECPKQ7zKPZLl4+ODO3fu4MCBAyWu8/jxY7P+7ZCwh0ZO2cFiFvn5+fj222+xbds2VKtWTRQNFsBmFmJGeYhXuUeyRo0ahZMnT2Lp0qV49OgRFAqFzuOXL1/GuXPn0KtXL4MXSUhJGjVqxHcJpBBrWaSmpqJatWpwdXXF0qVL+S7HpFjLQuwoD/Eq90iWpaUlgoKC4OXlhdDQUBw6dAgAMGfOHEyYMAHTpk2DhYUFpk2bZrRiCXlVYmIi3yWQQixl8dtvv2Hs2LFQq9UYN24c3+WYHEtZEMpDzCp08w5ra2usW7cOu3btQr9+/WBra4szZ87gn3/+QZMmTbBhwwZ07tzZWLUSUsxrr73GdwmkEAtZqNVqSKVSnDp1ClFRUXB0dOS7JF6wkAX5f5SHeFVqWh1PT094enoCAKRSKTiOo0tUCS+ys7PLvOqVmAbfWcTFxWHGjBnYvn071q1bx1sdLOA7C6KL8hCvKk8QXatWLUPUQUilWFtb810CKcRnFnl5eZg9ezZ++uknNG7cmLc6WEH7BVsoD/Gie/0TQgRLqVRiwYIFiI+Px6+//oo2bdrwXRIhhGhRk0UETalU8l0CKWTqLFQqFXx8fNClSxe0bt3a7KbGqQraL9hCeYhXlQ8XEsInOzs7vksghUyZxa+//gpXV1fs37+fzgfVg/YLtlAe4kUjWUTQ0tPT+S6BFDJFFmq1Gp9++in+/PNPtGjRghqsEtB+wRbKQ7xoJIsIWoMGDfgugRQydhZPnjyBjY0NJk2ahI4dOxr1tYSO9gu2UB7iRSNZRNDoJn/sMGYWhw8fxocffoj8/HxqsMqB9gu2UB7iRSNZRNBo4lV2GCOLvLw8KJVKSKVSREZGonr16gZ/DXNE+wVbKA/xopEsImg08So7DJ3Fw4cP4e3tjfv37+P999+nBqsCaL9gC+UhXjSSRQStbt26fJdAChkyC7Vajd27d+Pnn39G06ZNDfa8YkH7BVsoD/GikSwiaFlZWXyXQAoZIovc3FzMnDkTJ0+exHfffUcNViXRfsEWykO8aCSLCJqtrS3fJZBChshi3rx5GDlyJAYNGmSAisSL9gu2UB7iRU0WETSVSsV3CaRQVbIICQmBjY0NNm/eTHduNwDaL9hCeYgXHS4kgqZWq/kugRSqbBZffvklHj16BF9fX2qwDIT2C7ZQHuJFI1lE0GxsbPgugRSqaBZ37tzBs2fPsGjRItSqVctIVYkT7RdsoTzEi0ayiKBJpVK+SyCFKpJFSEgIvvrqK3Tu3JkaLCOg/cLw1q9fDxcXF/zzzz8V3pbyEC8aySKCRpdGs6M8WWRnZ+Phw4fw8PDA+PHjUa1aNRNUJj60Xxje3bt3YWFhAVdX1wpvS3mIF41kEUFLSkriuwRSqKwsbt68iREjRiA7Oxvt27enBsuIaL8wvHv37qF58+aws7Or8LaUh3gxOZKVnp6OTZs24ezZs0hNTUWLFi0wadIk+Pn5lWv7o0ePYs+ePXj27Bns7e0xcOBAfPLJJ6hTp47e1/njjz+QlJQEe3t7dO3aFR999BFat25tjLdGDIymq2BHSVlwHIfU1FQ8efIEBw4cQP369U1cmfjQfmFYycnJSE5ORpcuXSq1PeUhXsyNZMlkMgQEBCAsLAwDBw7EV199BUdHRyxatAhBQUFlbh8cHIwvv/wSDg4O+OKLLzB8+HAcOnQIkyZNQl5ennY9uVyOSZMmYd++ffD09MTXX38NPz8/XLhwAWPGjMHDhw+N+TaJgdB0FezQl0VWVhb8/f0RERGBkSNHUoNlIrRfGNbdu3cBAG3btsXff/+NSZMmoWPHjujatSu+/vpr5OTklLo95SFezI1khYaG4s6dO1i3bh2GDh0KABg7diymT5+OTZs2YcSIESX+ViCRSBAYGIhevXohODgYlpYFPaSbmxvmzZuHkJAQTJ8+HQCwe/duPHjwAMuWLcPYsWO1z+Hl5YUxY8bgxx9/xLZt24z8bklV0Q9tdryahVqtxoEDBzBr1iy88847PFUlTrRfGNa9e/cAANHR0QgKCsK7774LV1dX/Pnnnzh06BCkUik2bNhQ4vaUh3gxN5IVERGBhg0bahssALCwsMC0adOgVCoRFRVV4rZRUVFQKpWYMmWKtsECgGHDhqFJkyYIDw/XLvv7779RvXr1Yocg3dzc8MYbb+DatWsGfFfEWNLS0vgugRTSZMFxHDZt2oRly5bhgw8+oAaLB7RfGJZmJOvp06c4duwYVq1aha+++goRERFo1qwZTp06hcePH5e4PeUhXkw1WVKpFE+ePEGHDh2KPaZZduvWrRK3v3nzps66RbVv3x5PnjzRXkq7du1aHD58uNjJt5rzR4o2aYRd9vb2fJdACmmyCAoKgkKhwLfffstzReJF+4VhaUayli1bhsaNG2uX16xZEyNGjACAUk8xoTzEi6lOIjExERzH6T0caGtrizp16iAuLq7E7SUSCWrXrq33H3SjRo0AAPHx8QAKhm/1XYobFRWF5ORkdO3atbJvg5iQQqHguwRS6NKlS1i+fDlmzpyJzz77jO7eziPaLwwnOzsbsbGx6NChAzw9PYs9rjkUKJfLARRModOvXz/8/PPP2nUUCgVCQkLQsWNH3L9/3zSFEyYw1WRpRplq1qyp93EbGxvk5uaWun1p2wIFJ9aX5MGDB1i+fDmsrKwwe/bsUmtVKBSQSCRQq9XakxoTEhKgUCiQkpKCnJwcZGZmIj09Hbm5uUhKSkJ+fr7OuiqVCklJSZDJZMjIyEBmZiZkMhmSk5OhVCp11lWr1UhMTEReXh7S0tKQlZWF7OxspKSkQKFQ6KzLcRwSEhIgl8uRmpqK7OxsSKVSpKWlIS8vT2/dSqUSycnJFao7NzdXW3dOTk6F6k5NTdXWralX87dCodDWnZWVpa07MTGxWN35+fk6dWdkZGjrVqlUxdbV1J2enl5m3RKJRFu3VCrV1i2Xy3XqfTX78tSt+bxlMpm2bplMVqG69WWvqVsul2vrlkql2rolEkmZdaenp2vr1pe9pu6MjAxt3bt27cLmzZvh5+cHCwsLbd2JiYnaurOyskqsm+M4bd2pqakVqjsnJ0dbd25uboXqTk5O1vt5a/7NauoubV/TV3daWpq2bj6+IzRZ0ndE8bor+h1x/fp1cByHPn366N3XYmNjAQCvvfYasrOzkZmZiUmTJmHnzp3a74qLFy9i9erV+PHHH1GvXj1RfkeUtq+Z6jsiKysLJscx5J9//uGcnZ25devW6X28R48e3ODBg0vc/t133+V69uyp97F169Zxzs7O3L///qv38Tt37nDdunXjnJ2dudDQ0BJfIycnh4uOjuZycnJKeSfEVCgHfqWmpnLbtm3jpFIpl52dzXc5pBDtF4aze/duztnZmYuMjNT7+PDhwzkPDw9OLpdrl+Xm5nLdunXjwsLCuHv37nEdO3bk9u7da6qSSQn4+PnN1EiW5iZvRW+1UFReXl6pU3DY2dmVuK1mBEzfocRz585h4sSJSE9Px6JFizBhwoSKlk54Utal08R4oqOj4efnh06dOsHe3r7UUWJiWrRfGI7mfKzMzMxij505cwYxMTHw8/ODtbW1drmNjQ38/f0RHByMmTNnYvDgwfD39zdZzYQdTDVZTZs2hYWFBSQSSbHHZDIZsrKytOdWlbS9Zjj9VRKJBJaWlmjYsKHO8v3792PWrFlQKBRYs2YNJk2aVPU3QkzGwcGB7xJER61W46+//kK9evVw9OhRvPXWWwAoC5ZQFoajubIwKioKKpVKuzwuLg7fffcd6tevj1mzZhXbztfXF3FxcWjWrBkWL15ssnoJW5hqsuzs7NC6dWvcvn272GOaKwc7depU4vbu7u4A9F+BePv2bbRp00ZnJGv37t1YsmQJ7O3tsWvXLgwbNqyqb4GYWEpKCt8liEpKSgpGjx6NO3fuoEWLFjqzKFAW7KAsDEOhUODx48fo3Lkz4uPjMWbMGKxduxZfffUVvL29kZWVhU2bNsHR0VFnO7VajaVLl6JGjRrIyMhAeno6T++A8I2pJgsAvL29ER8fj+PHj2uXcRyHnTt3wtraGkOGDClxWy8vL1SvXh07duwAx3Ha5ceOHcPLly/h6+urXXbhwgWsWrUKDg4OCA0NrfR0CYRfNF2F6SQnJ+Pp06f45ptvMGPGjGKPUxbsoCwM48GDB1CpVOjYsSP27t0LBwcHhISE4OzZs+jbty/Cw8Ph4eFRbLuVK1fi/v37OHDgAJ4+fYqYmBjTF0+YYMEV7UYYkJeXh1GjRuH58+fw9/dHy5YtcfLkSVy8eBHz589HQEAAACA2NhbXr19Hs2bN0LFjR+32mzZtQmBgILp37w4vLy88ffoUISEhcHV1RWhoqPYqw8GDB+Pp06fw9fVFt27d9Naiuf9JUTKZDPfu3UPbtm1LvJKRmE5CQgL9QDEytVqN77//Hi9evEBwcHCJ61EW7KAs+BMSEoL169dj//79cHFxweLFi3Hnzh0cOXKE79JEj4+f38w1WUDB3XHXrVuHs2fPIicnBy1btsSUKVPg4+OjXSc8PBwLFy7EyJEjsWrVKp3t9+/fj9DQUDx//hz16tVD//798dFHH2kPbbx8+RJ9+/Ytsw59v31Qk8UWtVpNN441IrVajYsXLyImJgZTp04t9d5XlAU7KAt+/PHHH5g7dy62bNmCXr16ASgYEBg0aBD27t2Lzp0781yhuFGTJQDUZLElMTGx2MUMxDDOnj2LLVu24NChQ+W6sShlwQ7KwvTu3r2LCRMm4IsvvsD48eN1Hvvwww8hl8uxY8cOnqojAD8/v5mbIJqQiih64jUxnH/++Qe//PIL9uzZU+47t1MW7KAsTK9du3a4ceOG3sfWrl2rPVWFiAuNJxNBo3szGVZCQgJmzJgBDw8P7Ny5U3vvuvKgLNhBWbClrDzy84E//wT27y/4Oz/fJGURE6AmiwialRUNxhrKgwcP4O/vj08++aTYxOnlQVmwg7JgS2l5hIcDLVoAffsC48cX/N2iRcFyInzUZBFBo5N7q06lUuGnn35Cs2bNEBUVhbZt21bqeSgLdlAWbCkpj/BwwM8PiIvTXR4fX7CcGi3hoz2RCFpJ0yiR8klKSsKIESPQoEED2NjYwNbWttLPRVmwg7Jgi7488vOBjz8G9F16pln2ySd06FDoqMkigla7dm2+SxCsv//+GzVr1sSGDRuKXQ1VGZQFOygLtujL48KF4iNYRXEcEBtbsB4RLmqyiKClpqbyXYLgKJVKzJ8/H/v370f16tXxxhtvGOR5KQt2UBZs0ZdHQkL5ti3veoRNdHYkETS6q3XFJCUlwcrKCp6enhg5cqRBn5uyYAdlwRZ9eZQ3IopS2GgkiwhaAv2aV26RkZHw9/dHtWrVDN5gAZQFSygLtujLo2dPoGlToKTb0FlYAE5OBesR4aImiwhao0aN+C6BeWq1GhkZGbh48SIiIyONdqNKyoIdlAVb9OVRrRqwYUPBf7/aaGn+/6efCtYzqMzMkk8Gi4sreJwYDDVZRNAkEgnfJTDt6dOnGDJkCBQKBVatWoUaNWoY7bUoC3ZQFmwpKQ9fX+DwYaBJE93lTZsWLPf1NXAhmZnA4MFA794FZ9UXFRtbsHzwYGq0DIjOySKC5ujoyHcJzJLJZPj0008RFBSEBg0aGP31KAt2UBZsKS0PX19gxIiCqwgTEgrOwerZ0wgjWAAglQJJScCTJ0CfPgW3l3dyKmiw+vQpWK5Zj6ZmMggaySKClp2dzXcJzMnLy8Mnn3yClJQUHD16FC1atDDJ61IW7KAs2FJWHtWqFfQ448YV/G2UBgsoGCL780+gVav/b7QuXvz/BqtVq4LHmzY1UgHiQ00WETRjHv4SIoVCAR8fHwwYMADNmjUr9+TOhkBZsIOyYAtTeTg56TZa77yj22A5OfFdoVmhw4VE0Dh9t0sWqbCwMHTu3BmHDx+Gvb29yV+fsmAHZcEW5vJwcgJCQgoaLI2QEGqwjIBGsoigKZVKvkvgXX5+PmbNmoWbN2+iWbNmvDRYAGXBEsqCLczlERsL+PvrLvP3L34yPKkyarKIoNWsWZPvEngVExODpKQkzJ49GytXrkT16tV5q0XsWbCEsmALU3kUPcm9VSvg7791z9GiRsugqMkigpaRkcF3CbwJDQ3FvHnzAADt27fnuRpxZ8EayoItzOQRF1f8JPfu3YufDF/apIqkQuicLCJoprg1AWtkMhny8/OhVqsREREBKys2dmMxZsEqyoItzORRqxagqaXoSe6ak+H79Cl4vFYtngo0PzSSRQQtMTGR7xJM6s6dO/D29sajR48wadIkZhosQHxZsIyyYAszedSpA5w6BZw7V/wkdyenguWnTmnvkXXmzBm4uLggLCxMZ9W9e/fCxcUFgwcP1lmenp4Od3d3zJw506hvQ0jY+YYmpBLEMhEux3HgOA779u1DaGgok9OmiCULIaAs2MJUHnXqlHyj0Vfuj1W7dm0AQE5OjnaZWq3G3r17AQBZWVk66x86dAhyuRxTpkwxXL0CRyNZRNDEMBFudnY2pk6dit9//x3fffcdkw0WII4shIKyYItQ89DXZP3vf/9DbGws+vXrp9Nk5efnY//+/XB1dUW3bt1MXiurqMkiglavXj2+SzC6efPmYcqUKRg4cCDfpZRKDFkIBWXBFqHmoa/J2rVrFzp06ICePXtCqVRCJpMBAH7//Xe8fPmSRrFeQU0WETRmrtoxMI7jsG3bNhw9ehRBQUHo3bs33yWVyVyzECLKgi1CzePVJuvGjRu4ceMGpkyZglqFJ8drRrNCQkJQr149DB06FACgUqnQr18//PzzzzrPGRISgo4dO+L+/fumehu8oiaLCBpT958xoHnz5iEtLQ0jRoww6dQ4VWGuWQgRZcEWoeZhb28PKysrbZO1a9cuNGnSBIMGDdLe9DgjIwMPHjzA1atXMX78eFhbWwMArKysMHXqVOzatQsKhQIAcO7cOaxevRrr16+Hq6srP2/KxOjEdyJoKpWK7xIM6saNG0hISMDy5cthZ2fHdzkVYm5ZCBllwRYh51GrVi3k5OQgNjYWZ86cwfz581GtWjWdkaxjx46hRo0aGDdunM62fn5+2Lx5MyIiIuDu7o5PP/0UCxYsQJ8+fXh4J/ygJosImlqt5rsEg9m+fTtOnz6N4OBgwTVYgHllIXSUBVuEnEft2rWRk5ODPXv2wNbWFn5+fgCgHcmKjY1FZGQkvL294ejoqLOtjY0N/P39ERwcjPz8fIwcORL+r07nY+bocCERNBsbG75LqLKMjAzcuHED3bt3x8GDB1G3bl2+S6oUc8jCXFAWbBFyHrVr10ZCQgKOHDmC0aNHa5srzd+7d+9Gbm4uJk+erHd7X19fxMXFoVmzZvjqq69MVjcrqMkigiaVSvkuoUquXbsGX19fKJVKuLm5wdJSuLuk0LMwJ5QFW4ScR506dRAXFwe5XK4zCqVpsh48eIAePXqgTZs2xbZVq9VYunQpatSogYyMDFSrVs1kdbNCuN/ohACCHfXhOA5JSUmIj4/HkSNH8Pbbb/NdUpUJNQtzRFmwRch5aK4wfPfdd9GkSRPt8lpFpt4paRRr5cqVuH//Pg4cOICnT5/izz//NGqtLKImiwhaUlIS3yVUWFpaGsaOHYvffvsNPj4+eO211/guySCEmIW5oizYIuQ81q9fj5iYGPz00086y6tVq4aYmBjExMSgV69exbYLCQlBeHg4goKC0K5dO4wcORLBwcEmqpod1GQRQWNquopyyM/Px8GDBzF//nxMnDiR73IMSmhZmDPKgi1iy+OPP/7A6tWr8dNPP8HFxQUAMH36dNy8eRPR0dE8V2da1GQRQRPKdBVqtRo//vgjVq9ejZkzZ6Jz5858l2RwQslCDCgLtogpj7t37+Kzzz7DV199pTPC5eTkBC8vLwQFBfFYnelZcBzH8V2EkMhkMty7dw9t27YV7A3mzIlKpYKVFft3IgkMDISVlRVmzpwpmJuLVpRQshADyoItlAcb+Pj5TakTQUtLS0ODBg34LqNEf/31F86dO4dFixbxXYrRsZ6FmFAWbKE8xIuaLCJoRa9wYc2+fftw+vRpbNq0ie9STILlLMSGsmAL5SFedE4WETS5XM53CcUkJSVh27ZtGDFiBHbv3i2aL1gWsxAryoItlId4UZNFBI2185suXryIcePGwdPTE3Z2dszVZ0xieq+soyzYQnmIFx0uJILGysmk+fn5+Ouvv9C8eXP8+uuv2rshiwkrWRDKgjWUh3jRSBYRNJlMxncJSEhIwMiRI/Hs2TO0aNFClA0WwEYWpABlwRbKQ7yovSaC5uDgwOvrJyYmIjY2FqtWrUK7du14rYVvfGdB/h9lwRbKQ7xoJIsIWkpKCi+vq1Kp8PXXX+P777/H22+/LfoGC+AvC1IcZcEWykO8aCSLCBof01Vozr9q27YtJkyYYPLXZ5XYpg5hGWXBFspDvGgkiwiaqaerOHnyJCZMmIDevXtTg/UKMU0dwjrKgi2Uh3hRk0UErWHDhiZ7rUuXLuHYsWPYvXs3XZKthymzIKWjLNhCeYgXNVlE0JKTk43+Gi9evMAHH3yArl27YvPmzbCxsTH6awqRKbIg5UNZsIXyEC86J4sIWp06dYz6/Hfv3sUnn3yCrVu3wtKSficpjbGzIOVHWbCF8hAv+qlBBM1Y959RKBRYs2YNWrdujaioKLRu3door2NO6F5A7KAs2EJ5iBc1WUTQjHEn5ZcvX8Lb2xutW7dGjRo1UKNGDYO/hjmiu1qzg7JgC+UhXpQ8ETRDH8L7888/8dZbbyEoKAgtWrQw6HObOzqcyg7Kgi2Uh3hR8kTQDDW7vVwux0cffYRjx47BxsaGGqxKMFQWpOooC7ZQHuJFI1lE0GrVqlXl55BIJLCyssLgwYMxZMgQA1QlTobIghgGZcEWykO8aCSLCFpqamqVtj948CDef/992NjYUINVRVXNghgOZcEWykO8aCSLCFqjRo0qtV1+fj4yMjLw33//ITIyEtWrVzdwZeJT2SyI4VEWbKE8xItGsoigSSSSCm8TExMDLy8vWFhYYNmyZdRgGUhlsiDGQVmwhfIQL2qyiKBV9DdEqVSKr776Crt27YKjo6ORqhIn+m2dHZQFWygP8aImiwhaeX9DlMlkmDVrFqRSKY4cOYImTZoYuTLxod/W2UFZsIXyEC9qsoig1a1bt8x18vLy4OPjg1GjRqFx48YmqEqcypMFMQ3Kgi2Uh3hRk0UETSqVlvr4nj17IJFIEBERgQEDBpioKnEqKwtiOpQFWygP8WKyyUpPT8fy5cvRt29fuLu7w9vbG4cPHy739kePHoWPjw88PDzQo0cPLF26FJmZmaVuk5WVhV69eiEwMLCq5RMTKmnKG5VKhalTp+L58+dwcnJCzZo1TVyZ+ND0Q+ygLNhCeYgXc7dwkMlkCAgIwIMHDzB+/Hi0atUKp06dwqJFi5CSkoKZM2eWun1wcDDWrVsHT09PfPHFF4iLi0NISAiuX7+OsLAw2NjYFNsmNzcXs2fPRmJiorHeFjEStVpdbNnt27fRoEEDfP7552jXrh0PVYmTviwIPygLtlAe4sVckxUaGoo7d+5g3bp1GDp0KABg7NixmD59OjZt2oQRI0bg9ddf17utRCJBYGAgevXqheDgYO18UW5ubpg3bx5CQkIwffp0nW0ePHiAefPm4cGDB8Z9Y8QoVCqV9r85jsOOHTtw4sQJBAcHU4NlYkWzIPyiLNhCeYgXc4cLIyIi0LBhQ22DBQAWFhaYNm0alEoloqKiStw2KioKSqUSU6ZM0ZmQc9iwYWjSpAnCw8N11g8ODoaPjw8SExPx/vvvG/7NEKPTHAaUSqXIzs6GnZ0djhw5ggYNGvBcmfjQIVl2UBZsoTzEi6kmSyqV4smTJ+jQoUOxxzTLbt26VeL2N2/e1Fm3qPbt2+PJkyc6JyDevXsXvr6+OH78OPr27VvV8gkPMjMzcePGDYwYMQIvXrzA+PHjacZ7npR13iMxHcqCLZSHeDF1uDAxMREcx+k9HGhra4s6deogLi6uxO0lEglq164Ne3v7Yo9pbgYXHx8PV1dXAMCPP/4Ia2trAMCTJ08M8RaICXEcB0dHR2zcuBEHDx5EvXr1+C5J1OrXr893CaQQZcEWykO8mPqVXzPKVNLQqo2NDXJzc0vdvrRtgYIT6zU0DVZlKBQKSCQSqNVqJCQkAAASEhKgUCiQkpKCnJwcZGZmIj09Hbm5uUhKSkJ+fr7OuiqVCklJSZDJZMjIyEBmZiZkMhmSk5OhVCp11lWr1UhMTEReXh7S0tKQlZWF7OxspKSkQKFQ6KzLcRwSEhIgl8uRmpqK7OxsSKVSpKWlIS8vT2/dSqUSycnJFao7NzdXW3dOTk6F6k5NTdXWralX87dCodDWnZWVpa07MTFRW3dmZiZ8fX0RERGBTz75BLa2tsjMzERGRoa2bpVKpVNLfn6+tu709PQy65ZIJNq6NYcjU1NTIZfLdep9NfvS6n7185bJZNq6ZTJZherWl72mbrlcrq1bKpVq65ZIJGXWnZ6erq1bX/aaujMyMrR1JycnIyEhoVjdiYmJ2rqzsrJKrJvjOG3dqampFao7JydHW3dubm6F69b3eWv+zWrqLm1f01d3Wlqatm4+viOePXsm+u+Ikurm4zsiMTGRviNK2ddM9R2RlZUFU7PgOI4z+auW4Pr16xg3bhxmzpyJTz/9tNjjPXv2hL29PU6ePKl3+0GDBiE3Nxfnz58v9tj69esRFBSEgwcP6j2ceOXKFUyaNAlz5szB3LlzS6xRJpPh3r17aNu2LR1n59H06dMREBCAbt268V0KIYQQAeDj5zdTI1l2dnYACu7QrU9eXh5q1apV6vYlbasZAdN3KJEIA8dx2LBhA44dO4bt27ejW7du2t9aCP8oC3ZQFmyhPMSLqSaradOmsLCw0DvPk0wmQ1ZWVqkTbTZt2lQ7nP4qiUQCS0tLNGzY0KA1E9P58MMPAUDnylM6D4sdlAU7KAu2UB7ixVSTZWdnh9atW+P27dvFHtNcOdipU6cSt3d3dweg/wrE27dvo02bNjSSJUCXL1/GqVOnsGbNGnz88cewsLDQPpaRkcFfYUQHZcEOyoItlId4MdVkAYC3tzfi4+Nx/Phx7TKO47Bz505YW1tjyJAhJW7r5eWF6tWrY8eOHSh6qtmxY8fw8uVL+Pr6GrV2YniBgYHYuHEjPD099R5Dp/Pi2EFZsIOyYAvlIV5M3cIBACZPnozIyEgsWLAA//33H1q2bImTJ0/i4sWLmD9/vvYmk7Gxsbh+/TqaNWuGjh07AgCaNGmCmTNnIjAwEFOnToWXlxeePn2KkJAQtG/fHu+99x6fb41UQEpKCl68eIGBAwdizpw5OqNXRdGdlNlBWbCDsmAL5SFezI1k2djYICQkBD4+Pvj111/x3XffIT09HatXr0ZAQIB2vWvXrmH+/PkICwvT2X7OnDlYsmQJkpKSsGzZMpw8eRJjx47Fzp079c5bSNjz119/YcyYMbCwsICrq2uJDRYAMHRxrOhRFuygLNhCeYgXU7dwEAK6hYPxqNVqJCUl4cqVK+jXr1+pV5Jq5ObmwtbW1gTVkbJQFuygLNhCebBB9LdwIOKVlJQEX19fXLhwASNGjChXgwVAZ5okwi/Kgh2UBVsoD/Fi7pwsIj4qlQqHDx/GsmXLtFeIlpejo6ORqiIVRVmwg7JgC+UhXtRkEd7k5+fju+++g62tLb744otKPUdycrLeuS6J6VEW7KAs2EJ5iBc1WYQ3GzZsQLNmzTBlypRKPwd9cbGDsmAHZcEWykO8qMkiJnfmzBlcu3YNCxcurPJzJSQk0BcYIygLdlAWbKE8xIuaLGJSu3btwsWLF7FhwwaDPJ/mvmmEf5QFOygLtlAe4kVXFxKTiI+PR3BwMMaOHYvt27cb7PLZ1NRUgzwPqTrKgh2UBVsoD/GiJosY3Z9//onJkyejT58+Br83SXlv9UCMj7JgB2XBFspDvKjJIkajVCpx9uxZtGzZElFRUXBxcTH4a+Tl5Rn8OUnlUBbsoCzYQnmIFzVZxChevHgBb29vpKSkoHnz5ka727GlJf0TZgVlwQ7Kgi2Uh3jRie/E4OLj45GQkIDAwEC88cYbRn0tKyv6J8wKyoIdlAVbKA/xouSJwSgUCixcuBBWVlZYvXq1SV5TJpPBzs7OJK9FSkdZsIOyYAvlIV7UZBGDUKlUOH/+PN555x34+vqa7HUdHBxM9lqkdJQFOygLtlAe4kUHikmVhYeHY/LkyRgwYIBJGywASElJMenrkZJRFuygLNhCeYgXNVmkSs6fP4+///4bu3bt4uX16S7K7KAs2EFZsIXyEC9qskilPHr0CNOnT0fPnj2xdu1aWFtb81JHQkICL69LiqMs2EFZsIXyEC86J4tU2L///ouFCxciODgYFhYWvNbSsGFDXl+f/D/Kgh2UBVsoD/GikSxSbrm5ufj+++/Rtm1bREZGolmzZnyXhKSkJL5LIIUoC3ZQFmyhPMSLmixSLi9evMDw4cPRoUMH1KhRA9WrV+e7JAB01Q5LKAt2UBZsoTzEi5osUqbffvsNDg4O2L17N4YMGcJ3OTpkMhnfJZBClAU7KAu2UB7iRU0WKZFMJsP06dNx/vx51KxZE02bNuW7pGJYGVEjlAVLKAu2UB7iRSe+E73i4+NRvXp1jB07FgMGDOC7nBLxfeI9+X+UBTsoC7ZQHuJFI1mkmN27d2PWrFmoVasW0w0WAMjlcr5LIIUoC3ZQFmyhPMSLRrKIlkqlQmpqKuLj43H06FFUq1aN75LKZG9vz3cJpBBlwQ7Kgi2Uh3jRSBYBANy+fRuDBw+GjY0NFi1aJIgGCwDS0tL4LoEUoizYQVmwhfIQL2qyCDIyMrB8+XLs27cPderU4bucCmnUqBHfJZBClAU7KAu2UB7iRU2WiEmlUgQEBEChUODgwYNo0KAB3yVVmEQi4bsEUoiyYAdlwRbKQ7yoyRKpnJwc+Pj4YMqUKYJsrjRo4lV2UBbsoCzYQnmIFzVZIsNxHIKCgpCeno5jx46hZ8+efJdUJTTxKjsoC3ZQFmyhPMSLmiwRUSgUmDBhAqRSKRo3bgxbW1u+S6qyunXr8l0CKURZsIOyYAvlIV7UZInEP//8g4yMDCxZsgRffPEFLC3NI/qsrCy+SyCFKAt2UBZsoTzEyzx+0pIScRyHDRs2YNWqVbCysoKzszPfJRmUjY0N3yWQQpQFOygLtlAe4kU3IzVj6enpsLKywuuvv46DBw+a5dQOarWa7xJIIcqCHZQFWygP8aKRLDN1+fJljBo1CgkJCRgzZoxZNlhAwV3qCRsoC3ZQFmyhPMSLRrLMjFqthlqtxvHjxxEeHg4HBwe+SzKqmjVr8l0CKURZsIOyYAvlIV40kmVGUlNTMWbMGFy7dg3Lly83+wYLADIzM/kugRSiLNhBWbCF8hAvGskyI1988QW++uordOrUie9STKZ+/fp8l0AKURbsoCzYQnmIF41kCZxarcb333+P06dP4+effxZVgwUAiYmJfJdAClEW7KAs2EJ5iBc1WQL3wQcfoG7dunj33Xf5LoUXNF0FOygLdlAWbKE8xIsOFwrUuXPnIJfLsWnTJlHfgyUhIYG+wBhBWbCDsmAL5SFe1GQJ0OrVq3H//n0EBgaKusECgHr16vFdAilEWbCDsmAL5SFe1GQJiEQiwcuXL+Hr64s33njDbO99VREZGRl0UikjKAt2UBZsoTzEi87JEojff/8dEyZMQI0aNdCmTRtqsArZ2dnxXQIpRFmwg7JgC+UhXjSSxTiVSoWkpCTIZDJERkbSzvoKpVLJdwmkEGXBDsqCLZSHeNFIFsPi4+Ph4+OD6OhoDB8+nBosPTiO47sEUoiyYAdlwRbKQ7xoJItRSqUSERERWLNmDVxdXfkuh1nW1tZ8l0AKURbsoCzYQnmIFzVZjFEqlVi8eDEaN26MuXPn8l0O87Kzs2leMEZQFuygLNhCeYgXNVmMWb9+PTp06ID33nuP71IEwdHRke8SSCHKgh2UBVsoD/GiJosRx44dw927dzF//ny+SxGU5ORkuskfIygLdlAWbKE8xIuaLAYEBQXh7t27+OGHH/guRXDoi4sdlAU7KAu2UB7iRU0Wj549e4bTp09jypQpor9ze2XRdBXsoCzYQVmwhfIQL7qFA09Onz6NGTNmYODAgdRgVUGDBg34LoEUoizYQVmwhfIQL2qyTEwul+N///sf2rZti6ioKLRq1YrvkgQtNTWV7xJIIcqCHZQFWygP8aImy4QeP36M4cOHIzc3F82aNaN7pxhArVq1+C6BFKIs2EFZsIXyEC86J8tEYmNjkZycjO3bt6N58+Z8l2M28vLyYGtry3cZBJQFSygLtlAe4kVNlpHl5eVh3rx5qF+/PpYsWcJ3OWbH0pIGY1lBWbCDsmAL5SFe1GQZkVKpxLlz5zB06FAMGTKE73LMkpUV/RNmBWXBDsqCLZSHeFHyRvLLL7/gt99+w549e/guxazl5ubSxNmMoCzYQVmwhfIQLxrDNIIzZ87g7t272LlzJ9+lmL3atWvzXQIpRFmwg7JgC+UhXkw2Wenp6Vi+fDn69u0Ld3d3eHt74/Dhw+Xe/ujRo/Dx8YGHhwd69OiBpUuXIjMzs9h6+fn52L17N7y8vODu7o5+/fph/fr1yMvLq1Td9+7dwwcffID+/fvju+++oyFiE6BLo9lBWbCDsmAL5SFezHUBMpkMAQEBePDgAcaPH49WrVrh1KlTWLRoEVJSUjBz5sxStw8ODsa6devg6emJL774AnFxcQgJCcH169cRFhamc+PPpUuXIiwsDIMGDcKkSZNw9+5dBAcH47///sOOHTtgYWFR7rqvXbuGZcuWYdu2bRXajlQN3UWZHZQFOygLtlAeIsYxJjg4mHN2duaOHTumXaZWq7mAgADOzc2Ne/nyZYnbJiQkcG5ubty0adO4/Px87fKoqCjO2dmZ27Ztm3bZzZs3OWdnZ+6bb77ReY6goCDO2dmZO378uN7XyMnJ4aKjo7mcnByO4zguOzubW758OSeXyzmVSlWp90wqr7R/D8S0KAt2UBZsoTzY8OrPb1Ng7nBhREQEGjZsiKFDh2qXWVhYYNq0aVAqlYiKiipx26ioKCiVSkyZMkXnktlhw4ahSZMmCA8P1y7T/PfUqVN1nmPy5MmoUaOGzrolefLkCby9vdG9e3dYW1ujWrVq5X6fxDAaNmzIdwmkEGXBDsqCLZSHeDHVZEmlUjx58gQdOnQo9phm2a1bt0rc/ubNmzrrFtW+fXs8efIEUqlUu66DgwNatGihs56NjQ2cnZ1LfR0A+Ouvv1CvXj3s27cP/fr1K3VdYjxJSUl8l0AKURbsoCzYQnmIF1PnZCUmJoLjOL3Hr21tbVGnTh3ExcWVuL1EIkHt2rVhb29f7LFGjRoBAOLj4+Hq6gqJRFLicfJGjRrh9u3bkEqlxaZDUKvVAIC4uDhYWlqiVq1akMlk5X6PxLBsbGzo82cEZcEOyoItlAcbcnNzAfz/z3FTYKrJ0owy1axZU+/jNjY22g+ppO1L2xaA9h+6VCpFy5Yty1z31SZLLpcDKBgti4mJKbEWQgghhLBHLpfrHYwxBqaaLI7jdP7W93hZ0xOUtK1G0fOmSnudV9fVqFOnDlq0aIEaNWrQVAmEEEKIQKjVasjlctSpU8dkr8lUk6W5I25J96nKy8sr9VJYOzs7pKen631MMwKm6V7t7OxKfR1A/8zpVlZWqFu3bok1EEIIIYRNphrB0mBqKKZp06awsLCARCIp9phMJkNWVpb23KqSts/MzNR77FsikcDS0lJ7lUeTJk30vo5m3ddeew01atSo5DshhBBCiNgx1WTZ2dmhdevWuH37drHHNFcOdurUqcTt3d3dAei/AvH27dto06aNtovt0KED0tLSEBsbq7Nebm4uHjx4gI4dO1b6fRBCCCGEMNVkAYC3tzfi4+Nx/Phx7TKO47Bz505YW1tjyJAhJW7r5eWF6tWrY8eOHTrnWx07dgwvX76Er68vgIJpezTTHLz77rs60/bs3bsXCoVCu64+fE3bY45MNYVSUVlZWejVqxcCAwOrWr5ZMVUWmtfp168f3nzzTXTr1g0ff/wxHj9+bMi3I2imyiI7Oxtr167FoEGD4O7ujkGDBiE4OBhKpdKQb0fQ+PiOAoC1a9fCxcUFV65cqUr5ZsdUefz6669wcXHR++fLL78s9+tZcGWdKW5ieXl5GDVqFJ4/fw5/f3+0bNkSJ0+exMWLFzF//nwEBAQAAGJjY3H9+nU0a9ZMZ9Rp06ZNCAwMRPfu3eHl5YWnT58iJCQErq6uCA0NhVqtxsSJE/HgwQM0a9YMjx8/Rt26dZGamor27dvjv//+Q58+fbB161a90+MUnbZn4MCB2ml7WrduXWzansWLF2un7fH09MTdu3dx6NAhvPPOOxWetsccyWQybRZFp1C6dOkSPv300wpNoVRWFhq5ubmYPn06rl27hjlz5mDu3LnGenuCYqos5HI5/Pz88OjRI/j6+qJ9+/aIi4vDvn37YGFhgQMHDqBNmzameMvMMlUWmhs337hxA35+fmjXrh1u3LiBiIgI9OnTB8HBwaZ4u0zj4zsKAK5evYrJkydDrVZj79696Nq1qzHenuCYMo8ffvgBO3fuxIoVK2Btba3zPK/2HaUy2b3lKyA1NZVbtGgR5+npybm7u3MjRozgjh49qrPOkSNHOGdnZ27BggXFtt+3bx83ZMgQzs3Njevduze3bNkyLiMjg+M43Wl7lEolt2XLFq5///6ci4sL5+zszK1YsYLLzc3VW5cppu0RE1NNoaQRExPDDRs2jHN2duacnZ25jRs3GvYNCZipstD8+z9w4IDOc/z3339cu3btuOnTpxvwXQmTqbI4fPiw3n1l+fLlnLOzM/f3338b8F0Jk6m/oziO4zIzM7k+ffpwb775Jufs7MxdvnzZcG9I4EyZx9SpUzlPT88q18xkk2VMXl5eXM+ePYstv3TpEufs7MwFBweXuO22bds4Z2dn7q+//ir2WN++fbnBgwdr///bb7/lnJ2duadPn+qsl5uby7Vv354LCAio/JswE6bKguMKfri3bduW69KlC/f9999Tk/UKU2Xh7+/Pubm56Z3n09vbm/Pw8KjkOzAfpspi165d3LBhw7j09HSd9c6ePVtqEyAmpvyO0vj000+5Xr16cStWrKAm6xWmzOOdd97hJk2aVOWamTsny5iENG2PuTNlFgBw9+5d+Pr64vjx4+jbt29Vyzcrpsxi7dq1OHz4cLF70HEch9TUVNHfe86UWUyZMgVRUVFwcHDQWe/OnTsACq7AFjNTf0cBBXP3njhxAqtWrULt2rWrUr7ZMWUeqampSE5O1p66oFAooFAoKlW3qL7RTDVtj2bd0qbtyczMLLaDiYkpswCAH3/8EStWrED9+vUNUL15MWUW9evXh6ura7H1oqKikJycLPpzT0y9X2jI5XI8fvwYwcHBCAoKgru7OwYMGFCFdyJ8ps4iNjYWy5cvx+TJk+Hp6WmAd2BeTJnH/fv3AQAJCQnw9fWFh4cH3N3d4efnh0uXLlWoblE1Waaetqe864qRKbMAUOzERfL/TJ3Fqx48eIDly5fDysoKs2fPLm/ZZomvLMLCwjBkyBCsW7cOtWrVwpIlS0S/z5gyi/z8fMyfPx+vv/46Pvvss6qUbbZMmYdmyrx//vkHgwcPxqZNmzB//nwkJCQgICAAZ8+eLXfdTN3x3dg4AUzbIxamzoKUjM8s7t69i4CAAGRlZWHx4sV48803y1Gx+eIrCw8PD2zevBkSiQQ7duzA2LFjERgYKOpD66bMYuvWrbh9+zYOHTpEN8EugSnzcHd3x8yZM+Hr64vmzZtrHx80aBCGDRuGpUuXok+fPuU6vUFUI1nlmbZH31Q6RbcvaVtDTdsjFqbMgpSOryzOnTuHiRMnIj09HYsWLcKECRMqWrrZ4SsLzeHBiRMn4sCBA6hRowaWL19e0fLNiqmyuHnzJrZu3YopU6agYcOGSEtLQ1pamnYdqVSKtLQ0qNXqqrwdwTPlvtG5c2d8+umnOg0WUHCe4sCBAyGRSPDo0aNy1S2qJoum7WGHKbMgpeMji/3792PWrFlQKBRYs2YNJk2aVPU3YgZY2C8aNWqELl26ID4+vsS5YMXAVFmcP38eKpUK27dvh6enp/bPzp07AQAffvghPD098fLlS8O9OQFiYd8AoJ27OCcnp1x1i6rJoml72GHKLEjpTJ3F7t27sWTJEtjb22PXrl0YNmxYVd+C2TBlFjNmzEC/fv2gUqmKrZuTkwNLS0tRn5dlqix8fHywa9euYn9GjBgBAFiwYAF27dol+ot2TLlvzJ49GwMHDtQ78qWZmaJZs2blqltUTRZgmml7AGD48OEAgO3bt+s8R3mm7RELU2VBymaqLC5cuIBVq1bBwcEBoaGh6NKli3HekICZKovGjRsjPj4eR44c0XmO6OhoREdHo3v37tpDNGJliiycnJzQvXv3Yn+cnJwAAG5ubujevbuoj3xomGrfqF+/Pl68eIGwsDCd57hy5QrOnz+P3r17a0e0ysLctDrGZuxpe4reln/hwoUIDw/H4MGD0aNHD9y+fRsHDx4sddoeMTFlFkVduXIFkyZNoml1ijBVFoMHD8bTp0/h6+uLbt266a1F8xu8WJkqi9TUVIwZMwaJiYkYNWoU2rVrh4cPHyIsLAy1a9fGvn37ip2TIjZ8fUcBQGBgIDZt2kTT6hRhqjxSUlIwevRoJCYmwsfHB+7u7nj06BEOHDgAR0dH7Nu3D02bNi1f0VW+nakAGXPanqKKTtvj5ubG9e/fn/vpp59KnLZHjEyVRVGXL1+mO77rYews4uPjtVMalfaHmG6/SE1N5RYvXsz16NGDa9euHdezZ0/um2++4SQSibHemuDw8R3FcRy3ceNGuuO7HqbKIzk5mfvmm2+4nj17cu3ateN69OjBff311xXeN0Q3kkUIIYQQYgqiOyeLEEIIIcQUqMkihBBCCDECarIIIYQQQoyAmixCCCGEECOgJosQQgghxAioySKEEEIIMQJqsgghhBBCjICaLEIIIYQQI6AmixBCCCHECKjJIoQQQggxAiu+CyCEEFPLz89HeHg4jh49ipiYGCgUCjg7O2PWrFkYMGAA3+URQswEzV1ICBEVuVyOWbNm4e+//4abmxveeustyOVynD59GhkZGVi/fj2GDBnCd5mEEDNATRYhRFQ+++wzHD9+HIsXL8aECRO0yxMSEuDt7Q1HR0ecPn2axwoJIeaCzskihIjGpUuXcPz4cYwfP16nwQKA119/He+88w6ePXuG7OxsAMDixYvx+eef81EqIcQM0DlZhBDRCA0NBVBwTlZgYGCxx58/f67z/w8ePKBztAghlUZNFiFENC5evAgACAsLK3EdOzs72Nvbg+M4xMTE4MMPPzRVeYQQM0NNFiFEFLKysiCTydC/f39s2bKlzPVjY2Mhk8kgk8ng5+eHR48eoUOHDli9ejUaNWpkgooJIUJH52QRQkRBc41Penp6uda/f/8+LC0tERISgsWLF2Pfvn1IS0vDihUrjFkmIcSMUJNFCBGFOnXqoEWLFrh58yYuXbpU7HGFQoF///1X+/8xMTGoWbMmNm/eDHd3d7Rr1w4BAQG4du2aCasmhAgZHS4khIjGF198gblz52Lq1Kno3bs3WrdujdzcXEgkEkRHR2PgwIHw8PAAUDCSNWLECNSpU0e7va2tLeiuN4SQ8qImixAiGgMGDMDevXuxY8cO3LhxAxcuXICDgwMaNWqEMWPGwNfXV7tuTEwMevfurbP93bt34eLiYuqyCSECRU0WIURUunTpgi5dupS6TnZ2NuLi4pCfn69dlpeXh8OHD2PevHnGLpEQYiaoySKEkFc8ePAA1tbW2L9/Pzw8PFCtWjUsX74cLVu2xIgRI/gujxAiENRkEULIK+7fv4833ngD/v7+mDZtGvLy8jBs2DB8+eWXqFatGt/lEUIEguYuJIQQQggxArqFAyGEEEKIEVCTRQghhBBiBNRkEUIIIYQYATVZhBBCCCFGQE0WIYQQQogRUJNFCCGEEGIE1GQRQgghhBgBNVmEEEIIIUZATRYhhBBCiBFQk0UIIYQQYgTUZBFCCCGEGAE1WYQQQgghRvB/VFcgWcLJ+FAAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "# Set the font properties globally\n",
    "plt.rcParams['font.family'] = 'sans-serif'\n",
    "plt.rcParams['font.sans-serif'] = ['Arial']\n",
    "plt.rcParams['mathtext.fontset'] = 'dejavusans'\n",
    "\n",
    "\n",
    "# Define the range for x and y axes\n",
    "x_range = (0.0, 0.05)\n",
    "y_range = (0.0, 0.05)\n",
    "\n",
    "# Define the grid size\n",
    "grid_size = 0.01\n",
    "\n",
    "# Define the points\n",
    "b_X = [0.041, 0.018]\n",
    "w_X = [0.044, 0.017]\n",
    "\n",
    "# Create the figure and axis\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "# Set the x and y axis limits\n",
    "ax.set_xlim(x_range)\n",
    "ax.set_ylim(y_range)\n",
    "\n",
    "# Set the grid\n",
    "ax.set_xticks(np.arange(x_range[0], x_range[1] + grid_size, grid_size))\n",
    "ax.set_yticks(np.arange(y_range[0], y_range[1] + grid_size, grid_size))\n",
    "ax.grid(True, linestyle=':', linewidth=0.5)\n",
    "\n",
    "# Increase font size for tick labels\n",
    "ax.tick_params(axis='both', which='major', labelsize=14)\n",
    "\n",
    "# Plot the 45-degree line\n",
    "ax.plot([x_range[0], x_range[1]], [y_range[0], y_range[1]], linewidth=0.5, linestyle='--', color='black')\n",
    "\n",
    "# Plot the points\n",
    "ax.scatter(b_X[0], b_X[1], color='blue', label='Black-optimal point')\n",
    "ax.scatter(w_X[0], w_X[1], color='red', marker='x', label='White-optimal point')\n",
    "\n",
    "# Add labels for the points\n",
    "eps = 0.0005\n",
    "ax.text(b_X[0]-eps, b_X[1]+eps, r'$b_X$', fontsize=14, verticalalignment='bottom', horizontalalignment='right')\n",
    "ax.text(w_X[0]+eps, w_X[1]-eps, r'$w_X$', fontsize=14, verticalalignment='top', horizontalalignment='left')\n",
    "\n",
    "# Add a legend\n",
    "legend = ax.legend(frameon=True, facecolor='white', framealpha=1, fontsize=14)\n",
    "# Add labels to the axes\n",
    "ax.set_xlabel(r'$e_b$', fontsize=14)\n",
    "ax.set_ylabel(r'$e_w$', fontsize=14)\n",
    "\n",
    "# Display the plot\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hypothesis test\n",
    "\n",
    "* Table 1\n",
    "* See `code/q_test_science_logreg.py` and `code/q_test_science_rf.py` for details"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0, 0.0, 0.0, 0.0, 0.0]\n",
      "0.0\n"
     ]
    }
   ],
   "source": [
    "# logistic regression\n",
    "import pickle\n",
    "import numpy as np\n",
    "p_b_list = []\n",
    "\n",
    "for seed in [0,1,2,3,4]:\n",
    "    filename = f'../result/science/test_science_logreg_{seed}_None.pkl'\n",
    "    with open(filename, 'rb') as file:\n",
    "        res = pickle.load(file)\n",
    "\n",
    "    p_b = res['p_b']\n",
    "    p_b_list.append(p_b)\n",
    "\n",
    "print(p_b_list)\n",
    "print(np.median(p_b_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0, 0.0001, 0.0, 0.0, 0.0]\n",
      "0.0\n"
     ]
    }
   ],
   "source": [
    "# random forest\n",
    "import pickle\n",
    "import numpy as np\n",
    "p_b_list = []\n",
    "\n",
    "for seed in [0,1,2,3,4]:\n",
    "    filename = f'../result/science/test_science_{seed}_None.pkl'\n",
    "    with open(filename, 'rb') as file:\n",
    "        res = pickle.load(file)\n",
    "\n",
    "    p_b = res['p_b']\n",
    "    p_b_list.append(p_b)\n",
    "\n",
    "print(p_b_list)\n",
    "print(np.median(p_b_list))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Frontier estimation\n",
    "* Figure 10 (A) and (B)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of objectives: 72\n",
      "iteration: 1/72\n",
      "Set parameter Username\n",
      "Academic license - for non-commercial use only - expires 2025-11-18\n",
      "iteration: 2/72\n",
      "iteration: 3/72\n",
      "iteration: 4/72\n",
      "iteration: 5/72\n",
      "iteration: 6/72\n",
      "iteration: 7/72\n",
      "iteration: 8/72\n",
      "iteration: 9/72\n",
      "iteration: 10/72\n",
      "iteration: 11/72\n",
      "iteration: 12/72\n",
      "iteration: 13/72\n",
      "iteration: 14/72\n",
      "iteration: 15/72\n",
      "iteration: 16/72\n",
      "iteration: 17/72\n",
      "iteration: 18/72\n",
      "iteration: 19/72\n",
      "iteration: 20/72\n",
      "iteration: 21/72\n",
      "iteration: 22/72\n",
      "iteration: 23/72\n",
      "iteration: 24/72\n",
      "iteration: 25/72\n",
      "iteration: 26/72\n",
      "iteration: 27/72\n",
      "iteration: 28/72\n",
      "iteration: 29/72\n",
      "iteration: 30/72\n",
      "iteration: 31/72\n",
      "iteration: 32/72\n",
      "iteration: 33/72\n",
      "iteration: 34/72\n",
      "iteration: 35/72\n",
      "iteration: 36/72\n",
      "iteration: 37/72\n",
      "iteration: 38/72\n",
      "iteration: 39/72\n",
      "iteration: 40/72\n",
      "iteration: 41/72\n",
      "iteration: 42/72\n",
      "iteration: 43/72\n",
      "iteration: 44/72\n",
      "iteration: 45/72\n",
      "iteration: 46/72\n",
      "iteration: 47/72\n",
      "iteration: 48/72\n",
      "iteration: 49/72\n",
      "iteration: 50/72\n",
      "iteration: 51/72\n",
      "iteration: 52/72\n",
      "iteration: 53/72\n",
      "iteration: 54/72\n",
      "iteration: 55/72\n",
      "iteration: 56/72\n",
      "iteration: 57/72\n",
      "iteration: 58/72\n",
      "iteration: 59/72\n",
      "iteration: 60/72\n",
      "iteration: 61/72\n",
      "iteration: 62/72\n",
      "iteration: 63/72\n",
      "iteration: 64/72\n",
      "iteration: 65/72\n",
      "iteration: 66/72\n",
      "iteration: 67/72\n",
      "iteration: 68/72\n",
      "iteration: 69/72\n",
      "iteration: 70/72\n",
      "iteration: 71/72\n",
      "iteration: 72/72\n"
     ]
    }
   ],
   "source": [
    "seed = 1\n",
    "lmd = 0.001\n",
    "step_size = 5\n",
    "\n",
    "df = pd.read_csv('../data/df_health_risk.csv')\n",
    "Y_column = 'gagne_sum_t'\n",
    "G_column = 'dem_race_black'\n",
    "Y_columns = ['cost_t', 'gagne_sum_t']\n",
    "X_columns = list(set(df.columns) - set(Y_columns) - set([G_column]))\n",
    "X_columns.sort()\n",
    "\n",
    "y = df[Y_column].values\n",
    "thres_percentile = 97\n",
    "y_thres = np.percentile(y, thres_percentile)\n",
    "y_binary = (df['gagne_sum_t'].values > y_thres)*1\n",
    "df[Y_column] = y_binary\n",
    "\n",
    "angles_degrees = np.arange(0, 360, step_size)\n",
    "angles_radians = np.radians(angles_degrees)\n",
    "sin_values = np.sin(angles_radians)\n",
    "cos_values = np.cos(angles_radians)\n",
    "\n",
    "obj_list = []\n",
    "for i in range(cos_values.shape[0]):\n",
    "    obj_list.append((cos_values[i], sin_values[i]))\n",
    "\n",
    "res = CV_frontier(df, Y_column, G_column, X_columns, obj_list,\n",
    "                    n_splits=5, seed=seed, lmd=lmd, solver=cp.GUROBI,\n",
    "                    verbose=False, standardize=True)\n",
    "\n",
    "\n",
    "# save the results\n",
    "path = '../result/science/'\n",
    "filename = f'res_LP_science_{lmd}_{seed}.pkl'\n",
    "\n",
    "with open(path+filename, 'wb') as pickle_file:\n",
    "    pickle.dump(res, pickle_file)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0259944338862855,\n",
       " 0.06917532268233499,\n",
       " 0.04318088879604949,\n",
       " 0.018563236729471667,\n",
       " 0.37577611319077503)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "seed = 1\n",
    "lmd = 0.001\n",
    "pickle_file_path = f'../result/science/res_LP_science_{lmd}_{seed}.pkl'\n",
    "\n",
    "# Load the pickle file\n",
    "with open(pickle_file_path, 'rb') as pickle_file:\n",
    "    res = pickle.load(pickle_file)\n",
    "\n",
    "step_size = 5\n",
    "angles_degrees = np.arange(0, 360, step_size)\n",
    "angles_radians = np.radians(angles_degrees)\n",
    "sin_values = np.sin(angles_radians)\n",
    "cos_values = np.cos(angles_radians)\n",
    "\n",
    "obj_list = []\n",
    "for i in range(cos_values.shape[0]):\n",
    "    obj_list.append((cos_values[i], sin_values[i]))\n",
    "\n",
    "num_obj = len(obj_list)\n",
    "e_b_list = res['e_b_list']\n",
    "e_r_list = res['e_r_list']\n",
    "\n",
    "ind_b = np.argmin(e_b_list)\n",
    "ind_r = np.argmin(e_r_list)\n",
    "B_x = e_b_list[ind_b]\n",
    "B_y = e_r_list[ind_b]\n",
    "R_x = e_b_list[ind_r]\n",
    "R_y = e_r_list[ind_r]\n",
    "\n",
    "e_b_temp = 1.0\n",
    "idx = 0\n",
    "for i in range(num_obj):\n",
    "    if e_b_list[i] < e_r_list[i]:\n",
    "        if e_b_temp > e_b_list[i]:\n",
    "            e_b_temp = e_b_list[i]\n",
    "            idx = i\n",
    "e_r_temp = e_r_list[idx]\n",
    "\n",
    "slope = (e_r_temp - B_y) / (e_b_temp - B_x)\n",
    "F_x = (B_y - slope * B_x) / (1 - slope)\n",
    "F_y = F_x\n",
    "F_x - R_x, F_x, R_x, R_y, (F_x - B_x)/F_x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((0.04318088879604949, 0.018563236729471667),\n",
       " (0.04318088879604949, 0.018563236729471667),\n",
       " (0.06917532268233499, 0.06917532268233499))"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(B_x, B_y), (R_x, R_y), (F_x, F_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 0\n",
      "iteration: 1\n",
      "iteration: 2\n",
      "iteration: 3\n",
      "iteration: 4\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.06282993343603833, 0.02436941396987154)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res, y_b = no_info_utilitarian(df, Y_column, G_column, X_columns, seed=seed)\n",
    "u_x = res['e_b']\n",
    "u_y = res['e_r']\n",
    "u_x, u_y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 1000x1000 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "points = np.array(list(zip(e_b_list, e_r_list)))\n",
    "plt.axes().set_aspect('equal')\n",
    "\n",
    "ax = plt.gca()\n",
    "ax.tick_params(axis='x', which='major', pad=8)\n",
    "\n",
    "# Compute the convex hull\n",
    "hull = ConvexHull(points)\n",
    "\n",
    "# Plot the points\n",
    "# plt.plot(points[:,0], points[:,1], 'x')\n",
    "\n",
    "# Plot the convex hull\n",
    "for simplex in hull.simplices:\n",
    "    plt.plot(points[simplex, 0], points[simplex, 1], 'k-', linewidth=1)\n",
    "\n",
    "# Optionally, you can also fill the convex hull with a color\n",
    "plt.fill(points[hull.vertices,0], points[hull.vertices,1], 'lightblue', alpha=0.4)\n",
    "plt.plot([0, 1], [0, 1], color='gray', linestyle='--')\n",
    "\n",
    "\n",
    "# plt.text(B_x-0.01, B_y+0.01, r'$b_X$', fontsize=12, ha='center', va='center')\n",
    "plt.text(R_x+0.1, R_y-0.015, r'$b_X = w_X$', fontsize=12, ha='center', va='center')\n",
    "plt.plot(B_x, B_y, 'o', color='red')\n",
    "plt.plot(F_x, F_y, 'o', color='red')\n",
    "plt.text(F_x-0.01, F_y+0.04, r'$f_X$', fontsize=12, ha='center', va='center')\n",
    "\n",
    "plt.plot([B_x, F_x], [B_y, F_y], color='red', linewidth=2)  # Adjust linewidth for thickness\n",
    "\n",
    "\n",
    "# no info utilitarian\n",
    "# plt.plot(u_x, u_y, 'o', color='gray')\n",
    "# plt.text(u_x+0.03, u_y-0.03, r'$u_X$', fontsize=12, ha='center', va='center')\n",
    "\n",
    "plt.xticks(fontsize=12)\n",
    "plt.yticks(fontsize=12)\n",
    "plt.xlabel(r'$e_b$', size=14)\n",
    "plt.ylabel(r'$e_w$', size=14)\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.grid(linestyle = \"dashed\")\n",
    "plt.figure(figsize=(10,10))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 1000x1000 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "points = np.array(list(zip(e_b_list, e_r_list)))\n",
    "plt.axes().set_aspect('equal')\n",
    "\n",
    "ax = plt.gca()\n",
    "ax.tick_params(axis='x', which='major', pad=8)\n",
    "\n",
    "# Compute the convex hull\n",
    "hull = ConvexHull(points)\n",
    "\n",
    "# Plot the points\n",
    "# plt.plot(points[:,0], points[:,1], 'x')\n",
    "\n",
    "# Plot the convex hull\n",
    "for simplex in hull.simplices:\n",
    "    plt.plot(points[simplex, 0], points[simplex, 1], 'k-', linewidth=1)\n",
    "\n",
    "# Optionally, you can also fill the convex hull with a color\n",
    "plt.fill(points[hull.vertices,0], points[hull.vertices,1], 'lightblue', alpha=0.4)\n",
    "plt.plot([0, 1], [0, 1], color='gray', linestyle='--')\n",
    "# plt.legend()\n",
    "\n",
    "plt.xlim([0.03, 0.09])\n",
    "plt.ylim([0.01, 0.07])\n",
    "\n",
    "sw_corner_x = B_x\n",
    "sw_corner_y = R_y\n",
    "F_x_new = B_x\n",
    "F_y_new = B_x\n",
    "\n",
    "\n",
    "plt.plot([sw_corner_x, sw_corner_x], [sw_corner_y, 0.1], linestyle='-', color='blue', linewidth=1)\n",
    "plt.plot([sw_corner_x, 0.1], [sw_corner_y, sw_corner_y], linestyle='-', color='blue', linewidth=1)\n",
    "\n",
    "plt.text(R_x+0.002, R_y-0.003, r'$b_X = w_X$', fontsize=12, ha='center', va='center')\n",
    "plt.plot(R_x, R_y, 'o', color='red')\n",
    "# plt.text(B_x-0.003, B_y, r'$b_X$', fontsize=12, ha='center', va='center')\n",
    "plt.plot(B_x, B_y, 'o', color='red')\n",
    "plt.plot(F_x, F_y, 'o', color='red')\n",
    "plt.plot(F_x_new, F_y_new, 'o', color='blue')\n",
    "plt.text(F_x+0.002, F_y-0.002, r'$f_X$', fontsize=12, ha='center', va='center')\n",
    "plt.text(F_x_new+0.004, F_y_new-0.002, r\"$f_{X, G}$\", fontsize=12, ha='center', va='center')\n",
    "plt.text(0.08, 0.05, r\"$\\mathcal{E}_X$\",fontsize=16, ha='center', va='center')\n",
    "plt.text(0.047, 0.06, r\"$\\mathcal{E}_{X, G}$\",fontsize=16, ha='center', va='center', color='blue')\n",
    "\n",
    "plt.plot(u_x, u_y, 'o', color='gray')\n",
    "plt.text(u_x+0.002, u_y-0.002, r'$u_X$', fontsize=12, ha='center', va='center')\n",
    "\n",
    "# indifference curve\n",
    "x = np.linspace(0.03, 0.09, 100)\n",
    "y = u_y + (p_b/p_w)*(u_x - x)\n",
    "\n",
    "plt.plot(x,y, color='gray')\n",
    "\n",
    "plt.xticks(fontsize=12)\n",
    "plt.yticks(fontsize=12)\n",
    "plt.xlabel(r'$e_b$', size=14)\n",
    "plt.ylabel(r'$e_w$', size=14)\n",
    "\n",
    "# arrow\n",
    "plt.arrow(0.033, 0.027, -0.007*p_b/np.linalg.norm([p_b, p_w]), -0.007*p_w/np.linalg.norm([p_b, p_w]), color='gray', width=0.0003)\n",
    "plt.text(0.034, 0.022, r\"$H$\", color='gray', fontsize=16)\n",
    "\n",
    "# test\n",
    "# plt.plot(e_b_list[0], e_r_list[0], '+', color='black')\n",
    "# plt.plot(e_b_list[18], e_r_list[18], 'x', color='black')\n",
    "\n",
    "plt.plot([B_x, F_x], [B_y, F_y], color='red', linewidth=2)\n",
    "plt.plot([B_x, R_x], [B_y, R_y], color='red', linewidth=2)\n",
    "\n",
    "plt.grid(linestyle = \"dashed\")\n",
    "plt.figure(figsize=(10,10))\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.04269847486906801,\n",
       " 0.018564221535252855,\n",
       " 0.04469651312455836,\n",
       " 0.018148340945351005)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "B_x, B_y, R_x, R_y"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Robustness check\n",
    "* Appendix O.6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 48784 entries, 0 to 48783\n",
      "Columns: 149 entries, trig_min-normal_tm1 to nt_bnp_min-high_tm1\n",
      "dtypes: float64(15), int64(134)\n",
      "memory usage: 55.5 MB\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('../data/df_health_risk.csv')\n",
    "Y_column = 'gagne_sum_t'\n",
    "# G_column = 'dem_female'\n",
    "G_column = 'dem_race_black'\n",
    "Y_columns = ['cost_t', 'gagne_sum_t']\n",
    "X_columns = list(set(df.columns) - set(Y_columns) - set([G_column]))\n",
    "X_columns.sort()\n",
    "\n",
    "y = df[Y_column].values\n",
    "thres_percentile = 97\n",
    "y_thres = np.percentile(y, thres_percentile)\n",
    "y_binary = (df['gagne_sum_t'].values > y_thres)*1\n",
    "df[Y_column] = y_binary\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Point estimates with different algorithms\n",
    "* Try neural nets (NN) and support vector machine (SVM)\n",
    "* Are the point estimates close to the one by random forests?\n",
    "  - Previously, we have\n",
    "$b_X=(\\hat e_b^b, \\hat e_w^b) = (0.042, 0.019)$ and $w_X=(\\hat e_b^w, \\hat e_w^w) = (0.049, 0.018)$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NN (2-layer)\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.052, e_r=0.021, fairness=0.03\n",
      "[red] e_b=0.046, e_r=0.021, fairness=0.026\n",
      "r-skewed\n",
      "NN (3-layer)\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.054, e_r=0.021, fairness=0.033\n",
      "[red] e_b=0.048, e_r=0.021, fairness=0.027\n",
      "r-skewed\n",
      "SVM\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.057, e_r=0.022, fairness=0.034\n",
      "[red] e_b=0.062, e_r=0.024, fairness=0.038\n",
      "r-skewed\n"
     ]
    }
   ],
   "source": [
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "mlp = MLPClassifier(hidden_layer_sizes=(32, 32), max_iter=5000, activation='relu', solver='adam', random_state=42,\n",
    "                    early_stopping=True, validation_fraction=0.1, n_iter_no_change=30)\n",
    "print('NN (2-layer)')\n",
    "res_mlp = CV_classification(df, Y_column, G_column, X_columns, verbose=True, clf=mlp, standardize=True, ratio=None)\n",
    "\n",
    "\n",
    "mlp = MLPClassifier(hidden_layer_sizes=(32, 32, 32), max_iter=5000, activation='relu', solver='adam', random_state=42,\n",
    "                    early_stopping=True, validation_fraction=0.1, n_iter_no_change=30)\n",
    "print('NN (3-layer)')\n",
    "res_mlp = CV_classification(df, Y_column, G_column, X_columns, verbose=True, clf=mlp, standardize=True, ratio=None)\n",
    "\n",
    "print('SVM')\n",
    "svm = SVC(kernel='rbf', C=1.0, gamma='scale', random_state=42)\n",
    "res_svm = CV_classification(df, Y_column, G_column, X_columns, verbose=True, clf=svm, standardize=True, ratio=None)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Point estimates with different sample sizes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Estimate group-optimal points varying the size of the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train_frac: 0.7\n",
      "Logistic Regression\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.049, e_r=0.021, fairness=0.028\n",
      "[red] e_b=0.044, e_r=0.017, fairness=0.027\n",
      "r-skewed\n",
      "Random Forest\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.045, e_r=0.018, fairness=0.027\n",
      "[red] e_b=0.044, e_r=0.018, fairness=0.026\n",
      "r-skewed\n",
      "train_frac: 0.8\n",
      "Logistic Regression\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.048, e_r=0.021, fairness=0.027\n",
      "[red] e_b=0.042, e_r=0.017, fairness=0.025\n",
      "r-skewed\n",
      "Random Forest\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.042, e_r=0.019, fairness=0.024\n",
      "[red] e_b=0.043, e_r=0.018, fairness=0.026\n",
      "r-skewed\n",
      "train_frac: 0.9\n",
      "Logistic Regression\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.049, e_r=0.02, fairness=0.029\n",
      "[red] e_b=0.042, e_r=0.017, fairness=0.025\n",
      "r-skewed\n",
      "Random Forest\n",
      "\n",
      "blue: dem_race_black=1\n",
      "[blue] e_b=0.044, e_r=0.019, fairness=0.025\n",
      "[red] e_b=0.041, e_r=0.017, fairness=0.024\n",
      "r-skewed\n"
     ]
    }
   ],
   "source": [
    "for train_frac in [0.7, 0.8, 0.9]:\n",
    "    print(f'train_frac: {train_frac}')\n",
    "    clf_lr = LogisticRegression(max_iter=1000)\n",
    "\n",
    "    print('Logistic Regression')\n",
    "    res_logreg = CV_classification(df, Y_column, G_column, X_columns, verbose=True, clf=clf_lr, standardize=True, ratio=None, train_frac=train_frac)\n",
    "\n",
    "    clf_rf = RandomForestClassifier(n_estimators=500, random_state=42, n_jobs=-1)\n",
    "    print('Random Forest')\n",
    "    res_rf = CV_classification(df, Y_column, G_column, X_columns, verbose=True, clf=clf_rf, ratio=None, train_frac=train_frac)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}
